Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Transformations of cyclic olefins mediated by tungsten and molybdenum nitrosyl complexes Buschhaus, Miriam Sarah Anne 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2008_fall_buschhaus_miriam.pdf [ 13.24MB ]
Metadata
JSON: 24-1.0061738.json
JSON-LD: 24-1.0061738-ld.json
RDF/XML (Pretty): 24-1.0061738-rdf.xml
RDF/JSON: 24-1.0061738-rdf.json
Turtle: 24-1.0061738-turtle.txt
N-Triples: 24-1.0061738-rdf-ntriples.txt
Original Record: 24-1.0061738-source.json
Full Text
24-1.0061738-fulltext.txt
Citation
24-1.0061738.ris

Full Text

Transformations of Cyclic Olefins Mediated by Tungsten and Molybdenum Nitrosyl Complexes  by Miriam Sarah Anne Buschhaus B . S c , Trinity Western University, 2002  A THESIS SUBMITTED IN P A R T I A L F U L F I L L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF DOCTOR OF PHILOSOPHY  in T H E F A C U L T Y OF G R A D U A T E STUDIES (Chemistry) T H E UNIVERSITY OF BRITISH C O L U M B I A (Vancouver)  June 2008  © Miriam Sarali Anne Buschhaus, 2008  Abstract  Thermolysis ofCp*W(NO)(CH2CMe3)2,CpW(NO)(CH2CMe3)2,Cp*W(NO)(CH2SiMe3)(n2—CPhCH2),orCp*W(NO)[CH(Ph)CH2CHCPOCH2in cyclic olefins results in the formation of ring-retaining oligomers having lengths up to dodecamers. The main cyclohexene dimer is 3-cyclohexylcyclohexene. A small percentage of oligomers contain neopentyl or CH==CHPh end groups. Tumover frequencies for the Cp*-tungsten precatalysts range from 5.5 to 6.5 mol/h at 100 °C. In room temperature solutions,  Cp*Mo(NO)(CH2CMe3)2  generates the alkylidene  intermediate [Cp*Mo(NO)(=CHCMe3)], which couples with cyclic olefins to form cismetallacycles. The isolable cyclopentene-derived cis-metallacycle, Cp*Mo(NO)[cis-rn2 CH(CH2)3CHCHCMe3],  converts in the solid state to the allyl-hydride complex  Cp*Mo(NO)(H)(n3-CH(CH2)3CCHCMe3). With larger cyclic olefins (cyclohexene through cyclooctene) the initial cis-metallacycles isomerize to trans-metallacycles of the form Cp*Mo(NO)[trans-n2-CH(CH2)„CHCHCMe3] (n = 4, 5, 6), and these subsequently convert with loss of dihydrogen to n4-diene complexes, Cp*Mo(NO)[n4-CHCH(CH2)n-1CCHCMe3]. Thermolysis of the n4-diene complexes in cyclohexene results in decomposition of the organometallic complex with small amounts of oligomer formation. Thermolysis of Cp*W(NO)CH2CMe3)2 in cyclic-olefm substrates generates the alkylidene intermediate [Cp*W(N0)(=CHCMe3)], which couples with cyclic olefins in a manner analogous to the Cp*Mo-system. Tungsten trans-metallacycles are observed by*HN M R spectroscopy, but the organometallic subsequently reacts further with loss of the coupled neopentyl-cyclic olefin and coordination of two substrate molecules to form the putative Cp*W(NO)(cyclic olefm)2 complex. Two additional cyclooctene products are isolated, the 1,4diene  Cp*W(N0)[n4-CHCH(CH2)5CHCCH(CH2)6]  CH(CH2)6CCCHCH(CH2)5],  and the allyl hydride Cp*W(NO)(H)[n3-  both containing two coupled cyclooctene molecules. A tungsten  cis-metallacycle forms with 2,5-dihydrofiiran, but a ring-opened alkoxy-allyl complex forms with 3,4-dihydro-2H-pyran, and 1,2,3,6-tetrahydro-pyridine undergoes N - H bond activation to afford an amido product. CpW(NO)(CH2CMe3)2 produces some oligomers of cyclohexene, but in all other reactions bimetallic decomposition pathways predominate.  The proposed cyclic-olefm oligomerization mechanism begins with precatalyst initiation by coupling of the reactive alkylidene or n2-alkyne intermediates with a cychc olefin. The catalytic cycle consists of a Cp'W(NO)(cyclic olefm)2 species that couples the olefins to form a metallacyclopentane, foUowed by P-hydrogen activation and transfer to form a coordinated olefin. Subsequent incorporation of cyclic olefin may extend the oligomer length by a similar process, or the oligomer may dissociate and the Cp'W(NO)(cyclic olefin)2 complex may be regenerated. Catalyst decomposition is bimolecular.  Table of Contents Abstract  ii  Table of Contents  iv  ListofTabies  xi  List of Figures  xiii  List of Schemes  xvii  List of Abbreviations  xx  Acknowledgments  xxi  Co-Authorship Statement  xxiv  Chapter 1. The Cp'M(NO)(CH2CMe3)2 Complexes and Their Reactivity with Cyclic O l e f í n s  1  1.1 Introduction  2  1.2 The Cp'M(NO)(CH2CMe3)2 complexes  2  1.2.1 Typical C - H activation chemistry initiated by these complexes 1.3 General reaction pathways for alkylidene complexes with olefms 1.3.1 Cyclohexene as a polymerization substrate  3 7 10  1.4 Scope of this thesis  12  1.5 Thesis format  13  1.6 References  14  Chapter 2 . The Organic Products: Oligomers of Cyclohexene and 1,4-Cyclohexadiene  16  2.1 Introduction  17  2.2 Results and Discussion  17  2.2.1 Origin of the cyclohexene and 1,4-cycIohexadiene oligomers  17  2.2.2 Physical data for the cyclohexene and 1,4-cyclohexadiene oligomer mixtures . .  18  N M R spectroscopy  2.2.2.1  'H  2.2.2.2  Low-resolution El-mass spectrometry  18 19  2.2.3 Transfer dehydrogenation in the oligomerization systems  22  2.2.4 G C / M S analyses of the cyclohexene and 1,4-cyclohexadiene oligomer mixtures.  24  2.2.4.1 G C data for the cyclohexene oligomer mixture  24  2.2.4.2 M S data for the cyclohexene oligomer mixture  26  2.2.4.3 G C and M S data for the 1,4-cyclohexadiene oligomer mixture  29  2.2.5 Concentration effects in the Cp*W(NO)(CH2CMe3)2 system  29  2.2.6 Bulk separations of the cyclohexene oligomers  32  2.2.6.1 Unsuccessful attempts in the separation of the cyclohexene oligomers . . . .  32  2.2.6.2 Vacuum distillation of the cyclohexene oligomer mixture  33  2.2.7 Further characterization of the cyclohexene dimers and Identification of the major mono-unsaturated cyclohexene dimer 2.2.8 Other cyclic olefin substrates  34 39  2.2.9 Comparisons of the cyclohexene oligomerization products obtained by thermolysis of 1 and 5 with the products produced by other systems  40  2.3 Summary  41  2.4 Experimental Procedures  42  2.4.1 General methods  42  2.4.2 Oligomerization of cyclohexene and 1,4-cyclohexadiene with 1,2, 5 and 6 . . . .  42  2.4.3 Physical data for the bulk cyclohexene and 1,4-cyclohexadiene oligomer mixtures  43  2.4.4 G C / M S analyses of the cyclohexene and 1,4-cyclohexadiene oligomer mixtures 2.4.5 Monitoring of transfer dehydrogenation by  44 N M R spectroscopy  45  2.4.6 Bulk separation: Unsuccessful attempts in separation of the cyclohexene oligomers  45  2.4.7 Bulk separation: Vacuum distillation of the cyclohexene oligomer mixture . . . .  46  2.4.8 Identification of the major cyclohexene dimer  47  2.5 References  49  C h a p t e r 3. R e a c t i o n s o f Cp*Mo(NO)(CH2CMe3)2 w i t h Cyclic Olefins  50  3.1 Introduction  51  3.2 Results and Discussion  51  3.2.1 General reaction mechanism for the molybdenum system  51  3.2.2 Reactivity of 3 with cyclopentene: The cis-metallacycle 7  53  3.2.3 Further reactivity of 7: The allyl-hydride complex 8  55  3.2.4 Reactivity of 3 with cyclooctene: The trans-metallacycle 9  58  3.2.5 Further reactivity of 9: Formation of ri'^-diene 10 upon heating  60  3.2.6 Reactivity of 10: Trapping the ri^-diene complex 11 with pyridine  63  3.2.7 Reactivity of 10: Reléase of the coupled organic ligand 12  64  3.2.8 Sequential observation of metallacycles and an ri'^-diene complex: Reactivity with cycloheptene 3.2.9 Reactivity of 3 with cyclohexene: The ri'^-diene complex 16  65 67  3.2.10 Evidence for the metallacyclobutane intermedíate 17 in the cyclohexene reaction 3.2.11 Cyclohexene-derived organic products  69 72  3.2.12 Reactivity of 16 with cyclohexene and with mixed solvents under thermolysis conditions  73  3.2.13 Reactivity of the CpMo system with cyclohexene  74  3.2.14 Qualitative comparisons of the cyclic olefin substrates  74  3.2.15 Reactions of 3 with acyclic olefins  74  3.3 Summary  76  3.4 Experimental Procedures  77  3.4.1 General Methods  77  3.4.2 Preparationof?  77  3.4.3 Preparationof8  78  3.4.4 Preparationof9  78  3.4.5 Preparationof 10  79  3.4.6 Preparationof 11  79  3.4.7 Preparationof 12  80  3.4.8 Preparationof 13  80  3.4.9 Preparationof 14  81  3.4.10 Preparationof 15  81  3.4.11 Preparationof 16  82  3.4.12 Experimental evidence for 17  83  3.4.13 UV/vis experiment investigating the reaction of 3 with cyclohexene  84  3.4.14 Spectroscopic and G C / M S evidence for organic products of cyclohexene  85  3.4.15 Reactivity of 16 with cyclohexene under thermolysis conditions  86  3.4.16 Further thermal reactivity of 16 in cyclohexene and in mixed solvents  89  3.4.17 Preparationof 18  89  3.4.18 Preparationof 19  90  3.4.19 Preparation of 20  90  3.4.20 Preparation of 21 and characterization of 23  91  3.4.21 X-ray Crystallography  92  3.5 References  96  Chapter 4. Reactions of Cp'W(NO)(CH2CMe3)2 Complexes with Cyclic Olefíns and Other Substrates  97  4.1 Introduction  98  4.2 Results and Discussion  98  4.2.1 Reactivity of 1 with cyclohexene: Evidence for the metallacycle 24  98  4.2.2 Reactivity of 1 with cyclopentene: The cis-metallacycle 25  100  4.2.3 Reactivity of 1 with cyclooctene  101  4.2.3.1 The first cyclooctene product: Evidence for the trans-metallacycle 26 . . . .  102  4.2.3.2 The second cyclooctene product: The 1,4-diene 27  102  4.2.3.3 The third cyclooctene product: The allyl hydride 28  105  4.2.3.4 The organic products of cyclooctene  107  4.2.4 Reactivity of 1 with 4-methylcyclohexene  108  4.2.5 Comparison of tumover numbers for the cyclic-olefm substrates with 1  109  4.2.6 Reactivity of 1 with oxygen-containing cyclic olefins  110  4.2.6.1 Thermolysis of 1 in 2,5-dihydofiiran: The cis-metallacycle 29  110  4.2.6.2 Thermolysis of 1 in 2,3-dihydofuran  112  4.2.6.3 Thermolysis of 1 in 3,4-dihydro-2H-pyran: The alkoxy allyl 30 4.2.6.4 Comparisons to the reactivity of 5 with oxygen-containing cyclic olefins . . 4.2.7 Reactivity of 1 with nitrogen-containing cyclic olefins  113 114 116  4.2.7.1 Thermolysis of 1 in 1,2,3,6-tetrahydro-pyridine: The amido complex 33 . .  117  4.2.7.2 Thermolysis of 1 in 3-pyrroline  119  4.2.8 A brief summary of the isolable Cp* tungsten organometallic products  120  4.2.9 Reactivity of 2: Trapping the alkylidene intermedíate with trimethylphosphine..  120  4.2.9.1 The decomposition products of 2 in cyclohexane  121  4.2.9.2 Reactivity of 2 with other substrates  122  4.2.9.3 Reactivity of 2 with cyclic olefíns  123  4.2.9.4 Reactivity of 2 with oxygen-containing cyclic olefins  123  4.2.9.5 Reacüvity of 2 with nitrogen-containing cyclic olefins  124  4.3 Summary  125  4.4 Experimental Procedures  126  4.4.1 General Methods  126  4.4.2 Reactivity of 1 in cyclohexene: Identification of 24  127  4.4.3 Reactivity of 1 in cyclopentene: Preparation of 25  127  4.4.4 Reactivity of 1 in cyclooctene: Preparation of 27 and 28, Identification of 26 . . .  128  4.4.5 Reactivity of 1 in 4-methylcyclohexene  129  4.4.6 Reactivity of 1 in 2,5-dihydrofuran: Preparatíon of 29  130  4.4.7 Reactivity of 1 in 2,3-dihydrofuran  131  4.4.8 Reactivity of 1 in 3,4-dihydro-2H-pyran: Preparation of 30  131  4.4.9 Reactivity of 5 with oxygen-containing heterocycles: Preparatíon of 31 and 32 .  132  4.4.10 Reactivity of 1 in 1,2,3,6-tetrahydro-pyridine: Preparation of 33  132  4.4.11 Reactivity o f l in 3-pyrroline  133  4.4.12 Reactivity of 2 with trimethylphosphine in THF and in cyclohexene  133  4.4.13 Reactivity of 2 with trimethylphosphine in cyclohexane  134  4.4.14 Reactivity of 2 in cyclohexane: Formation of 36, 37, 38, 39 and 40  134  4.4.15 Thermolysis of36 and 37 inCeDe  135  4.4.16 Thermolysis of 2 in a range of other substrates  136  4.4.17 Thermolysis of 2 in the solid state  136  4.4.18 X-ray crystallography  136  4.5 References  141  Chapter 5. Mechanistic Insights into the Oligomerization of Cyclic Olefíns by the Tungsten Nitrosyl Precatalysts 5.1 Introduction  142 143  5.2 Results and Discussion  144  5.2.1 Comparisons between the tungsten and the molybdenum systems  144  5.2.2 Altérnate oligomerization systems: Precatalysts 5 and 6  147  5.2.3 The proposed Cp*W(NO)(cyclic olefm)2 reactive species  150  5.2.4 The proposed catalytic cycle for oligomerization of the cyclic olefms  152  5.2.5 The catalytic cycle applied to the dimerization of allylbenzene by 3  154  5.2.6 Other examples of cyclic olefins coupled in the coordination sphere of Cp*W(NO) complexes 5.2.7 Comparison to historical cyclohexene oligomerization reactions  155 164  5.2.8 Mechanistic proposals for the reaction of 1 and 5 with the oxygen-containing cyclic olefins 5.2.9 Decomposition of the catalyst: Concentration effects revisited  165 167  5.3 Summary  169  5.4 Experimental Procedures  170  5.4.1 General Methods  170  5.4.2 Preparationof41  170  5.4.3 ' H N M R spectrum of 41 in cyclohexene añer heating at 50 °C for 16 hours  170  5.4.4 X-ray Crystallography  171  5.5 References  173  Chapter 6. Thesis Summary and Future Directions  174  6.1 Thesis summary  175  6.1.1 Precatalyst initiation  175  6.1.2 The catalytic cycle  177  6.1.3 Catalyst decomposition and substrate limitations  179  6.1.4 Comparison of the Legzdins' alkyhdene complexes (Cp*M(N0)(=CHCMe3)) to known olefin metathesis alkylidene complexes 6.1.5 Significance and impact  179 180  6.2 Future directions  181  6.3 References  183  Appendíx A. Solid-State Molecular Structures Determined by X-Ray Crystallography A . l Introduction  184 185  A.2 Products of Cp*W(NO)(CH2CMe3)2 with cyclohexene in the presence of dihydrogen  186  A.3 The reactivity of several novel Cp*W(NO)(CH2CMe3)(allyl) complexes  189  A.4 Peroxide-induced nitrosyl insertion  224  A. 5 Examination of a potential catalytic cycle for C-C and C - 0 bond formation via C-H bond activation  228  A.6 Selectiveortho-activationofarylC-HbondsbyCp*W(NO)(CH2CMe3)2  233  A. 7 References  241  List of Tables Table 2.1 Tumover frequencies for precatalysts 1, 5 and 6 at concentrations of 0.01 M in the oligomerization of cyclohexene  18  Table 2.2 Tabulated mass spectral data for cyclohexene and 1,4-cyclohexadiene oligomers Table 2.3 Summary of G C and M S data for the cyclohexene oligomer families  19 29  Table 2.4 Concentration effects on the moles of cyclohexene converted per mole of tungsten  30  Table 2.5 The effect of initial concentration on the moles of cyclohexene converted per mole of timgsten precatalyst 1 at various reaction temperatures and reaction times.. Table 3.1 Diagnostic ' H - and '^C-NMR data for compounds 7-10 and 13-17  31 56  Table 3.2 Summary of the relative ratio of remaining 16 and oligomeric organic products at various reaction temperatures and times Table 3.3 X-ray Crystallographic Data for Complexes 7, 8, 9,10,11 and 16  88 94  Table 4.1 Comparison of the cyclic-olefm substrates oligomerized by precatalyst 1 based on tumover number (TON) and oligomer distribution  110  Table 4.2 Summary of substrates reacted with 2 and the number of products formed. . . . 1 2 2 Table 4.3 X-ray Crystallographic Data for Complexes 27, 28, 29, 31, 32 and 33  139  Table 5.1 X-ray Crystallographic Data for Complexes 39, 41 and 42  172  Table A. 1 X-ray Crystallographic Data for Complex A l  188  Table A.2 X-ray Crystallographic Data for Complexes A2, A3 and A4  193  Table A.3 X-ray Crystallographic Data for Complexes A5 and A6/A7  197  Table A.4 X-ray Crystallographic Data for Complexes A8 and A9  201  Table A.5 X-ray Crystallographic Data for Complexes AlO, A l l and A12  205  Table A.6 X-ray Crystallographic Data for Complexes A13 and A14/A15  210  Table A.7 X-ray Crystallographic Data for Complexes A l ó , A17 and A18  215  Table A.8 X-ray Crystallographic Data for Complexes A19, A20 and A21  220  Table A.9 X-ray Crystallographic Data for Complexes A22 and A23  223  Table A. 10 X-ray Crystallographic Data for Complexes A24 and A25  227  Table A. 11 X-ray Crystallographic Data for Complexes A26, A27 and A28  232  Table A. 12 X-ray Crystallographic Data for Complexes A29, A30 and A31  237  Table A. 13 X-ray Crystallographic Data for Complexes A32 and A33  240  List of Figures Figure 2.1 Low-resolution El-mass spectrum of the cyclohexene oligomers  20  Figure 2.2 Low-resolution El-mass spectrum of the 1,4-cyclohexadiene oligomers  21  Figure 2.3  N M R spectrum of the reaction mixtvire of 5 in neat cyclohexene after  24 h demonstrating the concomitant formation of cyclohexane with the cyclohexene oligomers  23  Figure 2.4 ' H N M R spectrum of the reaction mixture of 5 in neat 1,4-cyclohexadiene after 24 h demonstrating the concomitant formation of benzene with the 1,4-cyclohexadiene oHgomers  23  Figure 2.5 A typical G C trace for the mixture of cyclohexene oligomers  25  Figure 2.6 A typical G C mass spectrum of a cyclohexene dimer  26  Figure 2.7 A typical G C mass spectrum of a cyclohexene trimer  27  Figure 2.8 A typical G C mass spectrum of a cyclohexene tetramer  27  Figure 2.9 A typical G C mass spectrum of a cyclohexene pentamer  28  Figure 2.10 A typical G C mass spectrum of a cyclohexene dimer capped with a neopentyl end group Figure 2.11 Relative oligomer percentages obtained from selective solvation experiments.  28 32  Figure 2.12 Composition of the vacuum-distillation fractions of the cyclohexene oligomers  33  Figure 2.13 ' H N M R spectrum of the distilled cyclohexene dimer mixture  34  Figure 2.14 ^•'C N M R spectnmi of the distilled cyclohexene dimer mixture  35  Figure 2.15 ' H N M R spectrum of cyclohexylcyclohexane produced by the complete hydrogenation of the cyclohexene dimer mixture  36  Figure 2.16 '^C N M R spectrum of cyclohexylcyclohexane produced by the complete hydrogenation of the cyclohexene dimer mixture  36  Figure 2.17 *H N M R spectrum of cyclohexylcyclohexane and 3-cyclohexylcyclohexene produced by the partial hydrogenation of the cyclohexene dimer mixture  37  Figure 2.18 '^C N M R spectrum of cyclohexylcyclohexane and 3-cyclohexylcyclohexene produced by the partial hydrogenation of the cyclohexene dimer mixture Figure 2.19 Expansions showing the characteristic olefmic peaks of 3-cyclohexyl-  38  cyclohexene in (a) the  N M R spectrum and (b) the '^C N M R spectrum  38  Figure 3.1 Solid-state molecular structure of 7  54  Figure 3.2 Solid-state molecular structure of 8  57  Figure 3.3 Solid-state molecular structure of 9  59  Figure 3.4 Solid-state molecular structure of 10  61  Figure 3.5 Solid-state molecular structure of 11  64  Figure 3.6 Time lapse  N M R data for the reaction of 3 in cycloheptene. Selected  diagnostic signáis in the 2.7 to 5.0 ppm región, t = 0.1 to 312 h Figure 3.7 Solid-state molecular structure of 16  67 68  Figure 3.8 UV/vis data monitoring the reaction of 3 with cyclohexene, first day only (wavelengths 260-580 nm)  70  Figure 3.9 Expanded UV/vis data from Figure 3.8, fírst day only (wavelengths 390-510 nm)  71  Figure 3.10 The relative percentages of compounds 3,17 and 16 present in the reaction mixture  84  Figure 3.11 ' H N M R spectrum of the crude product mixture obtained from the reaction of 3 with cyclohexene at rt for 3 days  85  Figure 3.12 G C trace of isolated oligomeric products from the reaction of 3 with cyclohexene  89  Figure 3.13 ' H N M R spectrum of the crude product mixture obtained from the reaction of 16 in cyclohexene at 70 °C for 3 weeks Figure 4.1 Solid-state molecular structure of 27 (part A )  88 103  Figure 4.2 ' H N M R spectrum of the product mixture from thermolysis of 27 in neat cyclohexene at 70 °C for 24 h, demonstrating the formation of cyclohexene oligomers, the formation of 28 and the remaining presence of unreacted 27  105  Figure 4.3 Solid-state molecular structure of 28  106  Figure 4.4 Solid-state molecular structure of 29  112  Figure 4.5 Solid-state molecular structure of 31  115  Figure 4.6 Solid-state molecular structure of 32  116  Figure 4.7 Solid-state molecular structure of 33  118  Figure 5.1 Solid-state molecular structure of 41  148  Figure 5.2 ' H N M R spectrum of the reaction mixture of 41 in neat cyclohexene after  thermolysis at 50 °C for 16 h  149  Figure 5.3 Solid-state molecular structure of 43  156  Figure 5.4 Solid-state molecular structure of 44  158  Figure 5.5 The effect of variable initial concentrations of 1 in cyclohexene on catalyst activity (tumover number)  167  Figure A. 1 Solid-state molecular structure of A l  187  Figure A.2 Solid-state molecular structure of A2  190  Figure A.3 Solid-state molecular structure of A3  191  Figure A.4 Solid-state molecular structure of A4  192  Figure A.5 Solid-state molecular structure of A5 Part A  194  Figure A.6 Solid-state molecular structures of a) A6 and b) A7  195  Figure A. 7 Solid-state molecular structure of A6/A7, showing the crystallographic mirror-plane and the relative positions of Part A , Part B, and the shared methyl-allyl ligand  196  Figure A.8 Solid-state molecular structure of A8  199  Figure A.9 Solid-state molecular structure of A9  200  Figure A. 10 Solid-state molecular structure of AlO  202  Figure A . l l Solid-state molecular structure of A l 1  203  Figure A . 12 Solid-state molecular structure of A12  204  Figure A. 13 Solid-state molecular structure of A13  207  Figure A . 14 Solid-state molecular structure of A14  208  Figure A. 15 Solid-state molecular structure of AIS  209  Figure A . 16 Solid-state molecular structure of A16 Part B  212  Figure A . 17 Solid-state molecular structure of A17  213  Figure A . 18 Solid-state molecular structure of AIS  214  Figure A . 19 Solid-state molecular structure of A19  217  Figure A.20 Solid-state molecular structure of A20  218  Figure A.21 Solid-state molecular structure of A21  219  Figure A.22 Solid-state molecular structure of A22  221  Figure A.23 Solid-state molecular structure of A23  222  Figure A.24 Solid-state molecular structure of A24  225  Figure A.25 Solid-state molecular structure of A25  226  Figure A.26 Solid-state molecular structure of A26  229  Figure A.27 Solid-state molecular structure of A27  230  Figure A.28 Solid-state molecular structure of A28  231  Figure A.29 Solid-state molecular structure of A29  234  Figure A. 3 O Solid-state molecular structure of A30  235  Figure A.31 Solid-state molecular structure of A31  236  Figure A.32 Solid-state molecular structure of A32  238  Figure A.33 Solid-state molecular structure of A33  239  List of Schemes Scheme 1.1  3  Scheme 1.2  4  Scheme 1.3  5  Scheme 1.4  6  Scheme 1.5  7  Scheme 1.6  10  Scheme 2.1  17  Scheme 2.2  24  Scheme 2.3  39  Scheme 3.1  52  Scheme 3.2  53  Scheme 3.3  55  Scheme 3.4  58  Scheme 3.5  60  Scheme 3.6  63  Scheme 3.7  65  Scheme 3.8  66  Scheme 3.9  67  Scheme 3.10  75  Scheme 3.11  75  Scheme 4.1  99  Scheme 4.2  100  Scheme 4.3  102  Scheme 4.4  110  Scheme 4.5  111  Scheme 4.6  113  Scheme 4.7  114  Scheme 4.8  117  Scheme 4.9  117  Scheme 4.10  120  Scheme 5.1  143  Scheme 5.2  144  Scheme 5.3  145  Scheme 5.4  146  Scheme 5.5  147  Scheme 5.6  150  Scheme 5.7  151  Scheme 5.8  153  Scheme 5.9  154  Scheme 5.10  155  Scheme 5.11  157  Scheme 5.12  159  Scheme 5.13  160  Scheme 5.14  161  Scheme 5.15  162  Scheme 5.16  163  Scheme 5.17  165  Scheme 5.18  166  Scheme 5.19  167  Scheme 6.1  176  Scheme 6.2  177  Scheme 6.3  178  Scheme A . l  186  Scheme A.2  189  Scheme A.3  198  Scheme A.4  198  Scheme A.5  206  Scheme A.6  211  Scheme A.7  211  Scheme A.8  216  Scheme A.9  216 xviii  SchemeA.10 Scheme A . 11 SchemeA.12 Scheme A. 13 Scheme A. 14  '^'^^  List of Abbreviations a  alpha, herein the position once removed from the point of reference  P  beta, herein the position twice removed from the point of reference  Cp  cyclopentadienyl ligand, ri^-CsHs"  Cp'  Cp or Cp*  Cp*  pentamethylcyclopentadienyl hgand, T^^-CsMes'  A  heat  dn  deuterated in n positions  El  electrón impact  Et  ethyl, CH2CH3  GC/MS  gas chromatography mass spectrometry  T)  eta, herein denotes ligand hapticity  HOMO  highest occupied molecular orbital  L  ligand  LUMO  lowest unoccupied molecular orbital  M  metal, molar or Mega  MAO  methylalumoxane  Me  methyl, CH3  MS  mass spectrometry  V  nu, herein the stretching frequency, cm"'  NMR  nuclear magnetic resonance  NO  nitrosyl  ORTEP  Oak Ridge Thermal Ellipsoid Program  Ph  phenyl, CeHs  Pr  propyl, CH2CH2CH3  R  alkyl or aryl substituent or ligand  ROMP  ring-opening metathesis polymerization  rt  room temperature  'Bu  /er/-butyl, CMe3  THF  tetrahydrofuran, C4H8O  Acknowledgments Iwish to thank... The young woman at the desk paused and stared into space, her pen suspended motionless. Behind her, the other occupants of the room waited. Mará looked up from cleaning her fmgemails with her dagger tip and grumbled under her breath, "Isn't she fmished?" Her leather jerkin creaked as she shiñed position restlessly. Myric cast a sympathetic glance in Mara's direction. "She gave the thesis to him to read, so she must be almost done." "Good! Maybe we can fmally get some attention again. What's she writing now?" AlHon loosened his arm from around Myric's waist and peered over the shoulder of the writer. "The 'Acknowledgments.' But I think she has encountered writer's block again." Mará flourished her dagger to emphasize her point. "That should be easy. Loyal companions are the most important. You know, like all those research group members she's been working with for the past years. Jenkins and Peter G., lan, Chris, Scott... Craig and N e i l . . . Tve forgotten some of them." She scowled. Allion grinned. "What about their brave and fearless leader? She leamed so much from Dr. Legzdins. He gave her so many opportunities and so much guidance along the way. Not to mention those Latvian proverbs at lunch time. And Brian Patrick taught her to do X-ray crystallography." Gwennyth looked up as her spindle whirled. "Family is the most important." She glanced affectionately at Addiena and Erwyn playing on the floor with their adoptive father, his crippled feet out of harms way as they piled on his broad chest. "I know she loves her family a lot... her mother Maureen, her sisters Catherine and Anne, her brother Christopher and his wife Hannah. And her father, Detlev. He so wanted to live long enough to see her fínish her Ph.D. He must be proud of her, even though he has gone ahead of them now." Añer a moment of hush, Myric looked up at Allion and then across at Tenya whose dark head rested comfortably on blond Kierran's shoulder. "Yes, certainly family. And also friends who make it all worthwhile. She will miss the cióse ties she has at church. A l l her friends, younger and older." Gentle Mrym spoke softly, her palé face sad. "Sometimes it is enough to be thankful just to have fmished the task. But you have missed the most important of all. Him.''  "I wonder how long he will take to read it. And what he will think," Myric commented. "What's she wrhing now, my Heart's Own?" AUion leant over the writer again, scanning the page. Then he laughed aloud. "Everything we've been saying!" Exclamations of surprise broke out among the assembly. The young woman smiled, waiting for the clamor to die down. She raised her fmger-tips to brush against the velvety petáis of the crimson rose in the vase before her, as she thought of the one she loved most of all. Then the door swung open, and everyone fell suddenly silent. A silver circlet rested on the dark hair of the man who entered with a heavy sheaf of paper in his hand. The others drew back, some bowing with respect, as he came to the desk and laid down the tome. "I have read it." The young woman looked up, waiting. He smiled, and bent to kiss her forehead. "Well done." Then she smiled too, and picked up her pen to write the last words.  Co-Authorship Statement Portions of this thesis have been pubHshed and submitted for publication. Specifically, the chemistry described in Chapter 3 has been communicated (Graham, P. M . ; Buschhaus, M . S. A . ; Legzdins, P. J. Am. Chem. Soc. 2006,128, 9038-9039) and pubHshed as a full paper (Graham, P. M . ; Buschhaus, M . S. A . ; PampHn, C. B.; Legzdins, P. Organometallics, 2008, 27, 2840-2851). The chemistry described in Chapters 2, 4 and 5 has been submitted for publication (Buschhaus, M . S. A.; Pamplin, C. B.; Blackmore, I. J.; Legzdins, P. Transformations of Cyclic Olefms Mediated by Tungsten Nitrosyl Complexes). The author (Miriam Buschhaus) is responsible for all the research (including reactions, product characterization and data analysis) involving the compounds Cp'W(NO)(CH2CMe3)2 (Cp' = Cp or Cp*) with the cyclic olefms and the compounds Cp'Mo(NO)(CH2CMe3)2 with cyclohexene; for the characterization of the various organic cyclic-olefm oligomers; for the data coUection and analysis of all the solid-state molecular structures presented in this thesis; for the mechanistic proposals presented in this thesis; and for the writing of the entire thesis and of the ftiU paper based on Chapters 2, 4 and 5 currently submitted for publication. Dr. Peter Graham is responsible for the research (including reactions and product characterization) involving Cp*Mo(NO)(CH2CMe3)2 with cyclopentene, cycloheptene, and cyclooctene; and for the principal authorship of the two published papers mentioned above (but not for the v/riting of Chapter 3 of this thesis). Dr. lan Blackmore is responsible for the research (including reactions and product characterization of the organometallic complexes) involving Cp*W(NO)(CH2SiMe3)(r|'^CPhCHz) and Cp*W(NO)[CH(Ph)CH2CH("Pr)CH2] with cyclohexene. Dr. Craig Pamplin is responsible for initial exploratory reactions of cyclopentene with Cp*W(NO)(CH2CMe3)2 and Cp*Mo(NO)(CH2CMe3)2. Dr. Peter Legzdins is responsible for providing program oversight and discussion as the research supervisor.  C h a p t e r 1. T h e C p T V l ( N O ) ( C H 2 C M e 3 ) 2 C o m p l e x e s a n d Their Reactivity with Cyclic  Olefíns  1.1 Introduction This thesis describes the investigations into the reactivity of the title complexes with cyclic olefíns to form a series of novel organometallic compounds and oligomeric organic products, and the elucidation of the catalytic cycle by which the organic products are formed. This chapter introduces the title complexes and their reactive alkylidene intermediates, and briefly outlines the highlights of previous studies of the C-H bond activation chemistry initiated by these intermediates. A brief and general summary of known alkylidene reactivity with cyclic olefíns for a wide variety of transition-metal systems, particularly metathesis reactivity such as ring-opening metathesis polymerizatíon (ROMP), sets the broader context for the research presented in this thesis. At the end of this chapter the scope and organization of the thesis are briefly outlined.  1.2 The CpTVI(NO)(CH2CMe3)2 complexes Complexes of the general formula Cp'M(NO)(CH2CMe3)2 have been studied in the Legzdins group since the initial discovery of the fírst examples over two decades ago,^ and they have yielded a rich chemistry in C-H bond activation.'^"'' In these complexes, the central metal may be tungsten or molybdenum. The C p ' ligand may be cyclopentadienyl (Cp, ri^-CsHs) or pentamethylcyclopentadienyl (Cp*, ri^-CsMes). The general shape of the complexes is analogous to a three-legged piano stool, with the C p ' ligand forming the "seat" while the nitrosyl (NO) ligand and the two neopentyl (CH2CMe3) ligands form the three "legs". The geometry around the metal center is pseudooctahedral. Overall, the Cp'M(NO)(CH2CMe3)2 complexes have 16 valence electrons around the metal center, three of which are donated by the essentially linear nitrosyl ligand. Because of the influence of the nitrosyl ligand, the H O M O - L U M O gap is large. The L U M O is a nonbonding, metal-centered orbital located between the two neopentyl groups, and it gives the complexes Lewis acidic characteristics. Two methylene hydrogens, one fi"om each neopentyl group, form agostic interactions with the metal center in order to satisfy its need for more electrón density. The confluence of these factors makes the complexes prone to a-hydrogen abstraction.^ Thus, under appropriate reaction conditions, these complexes transfer an a-hydrogen from one neopentyl group to the other, resulting in a loss of neopentane and the formation of a reactive alkylidene intermediate which then initiates further chemistry (Scheme 1.1). For the  tungsten complexes, the alkylidene intermediates form under thermolysis conditions at 70 °C. In contrast, the molybdenum alkylidene intermediates form in solution at or below room temperature.^"'^  Scheme 1.1  Cp' M.^  Cp' - CMe4  RH O  M = Cp' = RH =  Cp' ,.M. O^ 7  -  WorMo C p * or Cp alkyl or aryl substrate with accessible C - H bond(s)  1.2,1 Typical C-H activation chemistry initiated by these complexes Previous work with the Cp'M(NO)(CH2CMe3)2 complexes has focused on the C-H bond activations initiated by the alkylidene intermedíate.^"'^ In the presence of an appropriate reagent with accessible aromatic or aliphatic hydrogens, the alkylidene reacts to actívate a C-H bond, reforming the neopentyl ligand by the reverse of the reaction by which the alkylidene initially formed and attaching the remainder of the reagent as a new aryl or alkyl ligand on the metal center (Scheme 1.1, step 2). Specifically, the typical single C-H bond activation reactivity of Cp*W(NO)(CH2CMe3)2 (1) is illustrated in Scheme 1.2. The reactive intermedíate, Cp*W(N0)(=CHCMe3), can be trapped with PMe3, with the alkylidene ligand being in two possible orientations. In deuterated benzene, 1 reacts to actívate a C-D bond. The activated deuterium is transferred stereospecifically to the a position on the neopentyl ligand, thus supporting the putative formation of an alkylidene intermedíate. As an example of single C - H bond activation of aliphatic bonds, tetramethylsilane reacts with 1 to yield products with trimethylsilylmethyl ligands.'^  Complex 1 can also effect múltiple C-H bond activations on one substrate (Scheme 1.3). Cyclohexane is activated twice, with complete loss of the alkylidene as neopentane, and the resulting cyclohexene adduct is trapped with PMes. Methylcyclohexane and ethylcyclohexane undergo triple C-H bond activation to form allyl-hydride products. These triple-activation products form even in the presence of PMe3, indicating that the C - H activation processes occur faster than the phosphine can trap any of the intermediate complexes.  The abiUty of complex 1 to selectively actívate specific C - H bonds varíes ín a mamier dependent on the class of substrate. The product distribution from the reaction with toluene indicates that the organometallic complex preferentially activates the stronger aryl C-H bonds over the alkyl C-H bonds^ and favors activation of the meta and para aryl bonds for steric reasons. In contrast, recent work has demonstrated that mono- and disubstituted benzenes with the formulae CeHsX and C 6 H 4 X 2 where X = Br, Cl, F, OMe and C s C P h are selectively orthoactivated by 1.^ Most aliphatic C-H bonds are activated indiscriminately, so that substrates such as «-pentane yield a plethora of C-H activated products, of which one has been identified as an  allyl hydride. Initial exploratory experiments also report that thermolysis of 1 in cyclohexene yields an "intractable mess", and since no organometallic products could be isolated, no further investigations of the cyclic-olefm substrates were pursued at that time. The related CpW(NO)(CH2CMe3)2 (2) complex has not been extensively studied ñor does the literature report any C-H activation reactivity for this complex, although the alkylidene intermedíate can be trapped.^ The molybdenum complex, Cp*Mo(NO)(CH2CMe3)2 (3), forms the reactive alkylidene intermedíate at ambient temperatures. The molybdenum alkylidene then activates a variety of CH bonds to form stable organometallic products in a manner similar to its tungsten analogue.'^"^ Activation of C^De results in stereospecific incorporation of a deuterium into the reformed neopentyl ligand. Activation of tetramethylsilane results in a single activation to yield only Cp*Mo(NO)(CH2CMe3)(CH2SiMe3). A variety of aryl substrates has also been activated.^ Overall, the reactivity of the molybdenum complex is analogous to that of tungsten, except that the reaction conditions are much milder and the products are slower to react further. The smaller Cp analog, CpMo(NO)(CH2CMe3)2 (4), forms an alkylidene intermedíate below room temperature and spontaneously dimerizes in CH2CI2 to yield an unsymmetrical complex with a unique bridging ri^ri'^-NO ligand (Scheme 1.4). The alkylidene also forms adducts with a variety of Lewis bases such as phosphines and pyridine, and activates a variety of heteroatom-hydrogen bonds to yield compounds of the general formula CpMo(NO)(CH2CMe3)(ER) where ER - NHR, OR, SCMe3, and OC(0)Me.''  Scheme 1.4  1.3 General reaction pathways for alkylidene complexes with olefins Since their original isolation in the 1970s, alkylidene complexes remain the focus of investigation due to their rich chemistry. To date, many isolable alkylidene complexes have been described in the literature.^ Others, such as those in the Legzdins systems, are proposed as reactive intermediates. Interest in alkylidene complexes that facilítate C-H bond activations continúes unabated.'^'^ Alkylidene complexes also have a rich chemistry of reactivity with olefins. In general terms, alkylidene complexes react with olefins according to one of the standard pathways illustrated in Scheme 1.5.'° The initial intermediate in all three cases is proposed to be a metallacyclobutane, formed by the addition of the olefin double bond to the M=C bond of the alkylidene. From that point the possible reaction pathways diverge to afford very different products, dependant on the transition metal and ligand-set combination employed in each system.  Scheme 1.5  R [M]  cyclopropanation  [M]  +  + R  R  olefin metathesis  /  R  H  [M]  \  R  R  P-H activation and olefin formation  Of the three reaction pathways, cyclopropanation  and p-hydrogen activation  are  relatively rarer. The allyl-hydride product formed directly by the p-hydrogen activation of the metallacyclobutane is often proposed as an intermedíate on the way to formation of an olefin, such as in a decomposition pathway proposed for one of the several Grubbs Ru-metathesis catalysts.'^ A stable allyl-hydride product has only recently been isolated in a (PNP)Ir=CH2 system (PNP = PPh2CH2SiMe2NSiMe2CH2PPh2)/° In contrast, olefin metathesis reactions are well known,*'''''' and the importance of these reactions was highlighted in 2005 with the awarding of the Noble Prize in Chemistry to Chauvin, Grubbs and Schrock for their contributions to the development of this field. The crucial involvement of alkylidene complexes in olefin metathesis is now well-understood, añer several decades of investigation. In some systems the initial alkylidene complexes may be isolated and fuUy characterized. In other systems the metallacyclobutane complex formed by reaction of the olefin with the alkylidene is isolable, while in many other systems it is an intermedíate and can not be isolated.*^'*'* The earliest olefin metathesis reactions, based on mixtures of tungsten or molybdenum chlorides with alkylaluminum cocatalysts, give no mechanistic insight, especially with regard to the catalytically active species. The first to propose the involvement of an alkylidene species in metathesis was Chauvin in 1971, at a time when only Fischer carbenes were known. The first well-defined nucleophilic alkylidene complex, Ta(CH2CMe3)3(=CHCMe3), was reported by Schrock in 1974.'^ Extensive work with the Tebbe complex, Cp2Ti(fa-CH2)(|a-Cl)AIMe2, elucidated the relationship between a metal alkylidene and a metallacyclobutane as applied to olefin metathesis, even though the complex itself does not readily catalyze metathesis reactions.'•^ Olefin metathesis reactions are now common, having rapidly developed and diversified from the initial discoveries to industrial applications, especially in polymer production and in complex organic syntheses.''' Two broad categories of metathesis catalysts exist: those based on earlier transition metáis (mainly tungsten and molybdenum) as exemplified by Schrock systems,''^''^ and those based on later transition metáis (mainly ruthenium) as exemplified by Grubbs systems.^''"'^ The latter are generally based on a RuCl2(phosphine)2(alkylidene) framework and on variations in which one phosphine ligand is replaced by an N-heterocyclic  carbene ligand.  Although often less reactive that their W and Mo counterparts, the Ru catalysts  show high tolerance for functional groups and for protic solvents, including water. The group-6 alkylidene complexes of Schrock'^ are of greater interest to us, as a comparison to our group-6 nitrosyl alkylidene compounds. The first d" tungsten alkylidene complex reported by Schrock was (PEt3)2Cl2W(0)(=CHCMe3), and it metathesizes olefin bonds slowly. The subsequently developed, fully active tungsten metathesis catalysts are built on a W(NAr)(=CHCMe3)(OR)2 motif where N A r is an imido group such as N-2,6-'Pr2C6H3 and (0R)2 represents two sterically bulky alkoxides (such as O'Bu), a biphenolate or a binaphtholate. A full range of molybdenum complexes of a similar type also exists. These pseudotetrahedral complexes are four-coordinate, 14 electrón, and sensitive to air and water. The alkylidene ligand takes one of two orientations relative to the imido ligand. The syn form, in which the 'Bu group of the alkylidene points towards the imido, generally predominates both in solution and in the solid State. The anti form ('Bu away from the imido) is significantiy more reactive to metathesis than the syn isomer. In the reaction with olefins, the olefin substrate coordinates to the metal center of the Schrock imido alkylidene to form an intermediate metallacyclobutane. Generally, the initial alkylidene complexes are well-characterized and subsequent alkylidene complexes can be detected by N M R spectroscopy, while the metallacyclobutanes are undetectable. A very few tungstenacyclobutanes are stable enough to isolate and characterize.'^ This offers a sharp contrast to the Legzdins system, where the alkylidene is a reactive intermediate undetectable by spectroscopy (although it can be trapped) and the metallocyclobutane is stable and fully characterizable (as described later in this thesis). The Schrock molybdenum catalysts offer greater metathesis reactivity than their tungsten analogues, oñen at lower temperatures, because the molybdenacyclobutane intermediate is more unstable.*'* Regardless of the metal employed, the catalysts are deactivated by undesired rearrangements of the metallocyclobutane intermediates to yield olefins (as in Scheme 1.5, P-H activation pathway) and by bimolecular decomposition, particularly if the neopentylidene ligand is changed to a methylidene in the process of the reaction.'^ In contrast, under the typical reaction conditions employed to genérate alkylidenes in the Legzdins system, the tungsten complex 1 is more reactive towards olefins than the molybdenum complex 3 (as demonstrated later in this thesis).  1.3.1 Cyclohexene as a polymerization substrate In general, when cyclic olefins are used as substrates in metathesis reactions the result is ring-opening to genérate linear oligomers or polymers by ROMP.*-''''* Scheme 1.6 illustrates a generic example of the first steps of this process for a simple cyclic olefin. The M=C bond of the alkylidene complex initially reacts with the cyclic olefin to form a metallacyclobutane, which then breaks apart to open the ring and regenérate a M=C linkage that subsequently incorporates another molecule of the cyclic olefin substrate. However, the cyclic olefin must have sufficient ring-strain to react in this manner.'"^ Thus, ring-opening metathesis polymerization of norbomene, cyclopentene, cyclooctene, and their derivatives is common, while similar processes for cyclohexene are unknown.  Scheme 1.6  Cyclohexene does not readily undergo ring-opening metathesis polymerization due to the relative lack of ring-strain in the monomer. Thermodynamically, the valué of AG° for the polymerization of cyclohexene is positive under all typical polymerization conditions and temperatures. The contribution of AH° cannot offset the large positive valué of-TAS°.*^''^ Some small amounts of ROMP-derived cyclohexene oligomers can be observed if the metathesis reaction is carried out at very low temperatures with a tungsten hexachloride / tetramethyltin catalyst system, but the oligomers revert to monomer at room temperature in the presence of the catalyst.*^ A stoichiometric 1:1 ring-opening of cyclohexene has been achieved with a Ru(PCy3)2Cl2(=CHC02R) (R = Me,;7-tolyl, 'Bu, 'Pr, cyclohexyl, 1-adamantyl) catalyst system,^"" and modifications to this system can catalytically open cyclohexene as linear RC(0)CH=CH(CH2)4CH=CHC(0)R (R - Et, O'Bu, OH) by ring-opening cross metathesis.^"" In this case the nature of the alkylidene complex drives the reactivity, overriding the usual unreactivity of the cyclohexene substrate.  Early attempts to polymerize cyclohexene with typical metathesis catalyst systems in the 1970s and 80s did not yield ring-opened products. Instead, the isolated polymers and oligomers of cyclohexene were reported to be ring-retaining. Thus, a Re(C0)5Cl + E t A l C l i system, heated at 110 °C for 24 h, yields ñilly saturated 1,2-polycyclohexene with an average molecular weight of 2500.^' Two metathesis systems based on tungsten hexachloride (WCle + Me4Sn at 50 °C for 7 days^^ and W C U + E t A l C l i + R O H (R = Et, Ph, PhCHz) at 70 °C for 24 h^^) are reported to yield short, ring-retaining oligomers of cyclohexene with almost complete saturation. The presence of dimers and trimers in these product mixtures is reported, but the exact identity of the isomers and the locations of the unsaturations have not been determined. The actual catalytically-active species carmot be identified in the reaction "soups" of these systems. Therefore, no mechanistíc insight into the organometallic species is reported, although a tungsten hydride is proposed as a possible species in the second system.'^^ Recently, cyclohexene has been successfuUy copolymerized with ethylene by titanium/MAO catalysts, the best incorporation being 16.2 mol% inclusión of cyclohexene using the catalyst C p T i C l 2 ( 0 A r ) (Cp' = l,2,4-Me3C5H2, OAr = 0-2,6-'Pr2C6H3).^^ The six-membered ring is retained in the structure of the polymer in a manner that suggests a 1,2-insertion mechanism, although no specific mechanistic studies are reported. The authors note that the absence of any cyclohexene repeat units suggests that incorporation of a higher percentage of cyclohexene may be difficult. The resistance of cyclohexene to typical polymerization methods has resulted in the use of 1,3-cyclohexadiene as a replacement substrate. 1,3-Cyclohexadiene can be polymerized by various methods, including systems based on radical, anionic, cationic, transition-metal or rareearth metal catalysts.'^^ Past work has focused on finding effective catalysts that are not poisoned by the 1,3-cyclohexadiene substrate, while recent work focuses on improving the regio- and stereoselectivity of the polymerization reactions.^^"^^ One recent report features cationic polynuclear rare-earth polyhydrido complexes that afford cis-l,4-linked poly(cyclohexadiene) with almost perfect selectivity."^^ Another report examines the effective homo- and copolymerization (with ethylene, 12.3 mol% inclusión of the diene) of 1,3-cyclohexadiene by the titanium catalyst system [Me2Si(N'Bu)(Me4C5)]TiCl2/MAO to form soluble polymers.^^ In both cases, the ring-motif is retained as a 1,4-cyclohexene unit. The unsaturation remaining in the polymer offers potential for two further reactions: hydrogenatíon to yield saturated polymers or functíonalization of the double bond. Hydrogenation of any 1,3-cyclohexadiene polymer  yields products that often called "polycyclohexenes", although cyclohexene is not the polymerization substrate.  1.4 Scope of this thesis This thesis explores the reactivity of the Cp'M(NO)(CH2CMe3)2 complexes with cyclic olefms, with particular emphasis on cyclohexene. The oligomeric organic products and the isolable organometallic compounds are described, and a catalytic cycle is proposed to rationalize the formation of the observed products. Chapter 2 describes the ring-retaining oligomers of cyclohexene and 1,4-cyclohexadiene obtained from thermolysis reactions of a variety of organometallic complexes, including 1 and 2, in the cyclic-olefm substrate. A transfer-dehydrogenation reaction is described, accounting for the increased levéis of unsaturation in the observed products. The structure of one monounsaturated dimer isomer is determined, and the absence of any R O M P products is noted. Chapter 3 outlines the reactivity of the molybdenum complexes, primarily Cp*Mo(NO)(CH2CMe3)2 (3), with the cyclic olefins ranging from cyclopentene to cyclooctene. A reaction pathway describing the key steps in the transformation from an initially formed cismetallacyclobutane complex to a final ri^'-diene complex is described. The limited ability of the molybdenum system to oligomerize cyclic olefins is explored. Chapter 4 presents the isolable organometallic products derived from the tungsten complexes. The main part of the chapter focuses on the reactivity of Cp*W(NO)(CH2CMe3)2 (1), and describes the characterization of the products that form under thermolysis conditions with cyclopentene, cyclooctene, and a variety of other cyclic olefins, including substrates that contain oxygen and nitrogen atoms within the ring. The last part of the chapter contrasts the reactivity of CpW(NO)(CH2CMe3)2 (2) and its propensity to decompose in most substrates, both with the cyclic olefms and other substrates with potential for C-H bond activation. Chapter 5 considers the insights that can be drawn from the total available data. Comparisons are drawn between the tungsten and the molybdenum systems, and a catalytic cycle is proposed which rationalizes both initiation and propagation for the cyclic olefin oligomerization observed in the tungsten system. The diversity of the organic products and the data suggesting possible catalyst deactivation trends are considered.  1.5 Thesis format This thesis contains six chapters and an appendix. Chapters 2 through 5 are each divided into five main sections: Introduction, Results and Discussion, Summary, Experimental Procedures, and References. Chapter 6 provides a brief summary of the thesis and discusses further directions. Finally, the appendix contains a summary of many other solid-state molecular structures determined by the author in addition to all those presented in the chapters, along with brief explanations of their importance to projects outside the main scope of this thesis. Compounds within the thesis chapters are numbered sequentially throughout as 1, 2,3, etc., while those appearing in the appendix are numbered sequentially as A l , A2, A3, etc.  1,6 References (1)  Legzdins, P.; Rettig, S. J.; Sánchez, L.; Bursten, B. E.; Gatter, M . J. J. Am. Chem. Soc. 1985,707, 1411-1413.  (2)  Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, 223-233 and references cited therein.  (3)  Legzdins, P.; Pamplin, C. B. Sequential Hydrocarbon C - H Bond Activations by 16electron Organometallic Complexes of Molybdenum and Tungsten. In Activation and Functionalization of C-H Bonds; Goldberg, K . I.; Goldman, A . S., Eds.; A C S Symposium Series 885; American Chemical Society: Washington, D.C., 2004; pp 184197.  (4)  Blackmore, I. J.; Jin, X . ; Legzdins, P. Organometallics 2005, 24, 4088-4098 and references cited therein.  (5)  Tsang, J. Y . K.; Buschhaus, M . S. A . ; Legzdins, P.; Patrick, B. O. Organometallics 2006, 25, 4215-4225.  (6)  Tran, E.; Legzdins, P., unpublished observations.  (7)  Wada, K.; Pamplin, C. B.; Legzdins, P.; Patrick, B. O.; Tsyba, I.; Bau, R. J. Am. Chem. Soc. 2003,125, 7035-7048.  (8)  Schrock, R. R. Chem. Rev. 2002,102, 145-179 and references cited therein.  (9)  a) Basuli, F.; Bailey, B. C ; Huffman, J. C ; Mindiola, D. J. Organometallics 2005, 24, 3321-3334 and references cited therein. b) Cheon, J.; Rogers, D. M . ; Girolami, G. S. J. Am. Chem. Soc. 1997,119, 6804-6813. c) Coles, M . P.; Gibson, V . C ; Clegg, W.; Elsegood, M . R. J.; Porrelli, P. A . J. Chem. Soc, Chem. Commun. 1996,16, 1963-1964. d) van der Heijden, H . ; Hessen, B. J. Chem. Soc, Chem. Commun. 1995, 2, 145-146.  (10)  Fryzuk, M . D.; Gao, X . ; Rettig, S. J. J. Am. Chem. Soc 1995, / / 7 , 3106-3117 and references cited therein.  (11)  Doyle, M . P. Chem. Rev. 1986, 86, 919-930 and references cited therein.  (12)  Janse van Rensberg, W.; Steynberg, P. J.; Meyer, W. H.; Kirk, M . M . ; Forman, G. S. J. Am. Chem. Soc 2004,126, 14332-14333.  (13)  Olefin Metathesis and Metathesis Polymerization; Ivin, K . J.; Mol, J. C ; Academic Press, San Diego, 1997 and references cited therein.  (14)  Handbook of Metathesis; Grubbs, R. H., Ed.; Wiley-VCH, Weinheim, 2003, vols. 1-3: Catalyst Development (vol. 1); Applications in Organic Synthesis (vol. 2); Applications  in Polymer Synthesis (vol. 3) and references cited therein. (15)  Schrock, R. R. J. Am. Chem. Soc. 1974, 96, 6796-6797.  (16)  a) Schrock, R. R.; Czekelius, C. Adv. Synth. Catal. 2007, 349, 55-77 and references cited therein. b) Schrock, R. R. J. Mol. Catal. A: Chem. 2004, 275, 21-30 and references cited therein.  (17)  Grubbs, R. H . Tetrahedron 2004, (50, 7117-7140 and references cited therein.  (18)  Olefin Metathesis and Ring-Opening Polymerization of Cyclo-Olefins, V . Dragutan, A . T. Balaban and M . Dimonie, Wiley-Interscience, Bucharest, 1985.  (19)  Patton, P. A . ; Lillya, C. P.; McCarthy, T. J. Macromolecules 1986,19, 1266-8.  (20)  a) Ulman, M . ; Belderrain, T. R.; Grubbs, R. H. Tetrahedron Lett. 2000, 41, 4689-4693. b) Choi, T.; Lee, C. W.; Chatterjee, A . K.; Grubbs, R. H . J. Am. Chem. Soc. 2001,123, 10417-10418. c) For similar reactions, see also: Lee, C. W.; Choi, T.; Grubbs, R. H. J. Am. Chem. Soc. 2002,124, 3224-3225, and Randl, S.; Connon, S. J.; Blechert, S. Chem. Commun. 2001, 1796-1797.  (21)  a) Farona, M . F.; Tsonis, C. P. J. Chem. Soc, Chem. Commun. 1977, 363-364. b) Wazeer, M . I. M . ; Tsonis, C. P. Polym. Bull 1984,12, 475-479.  (22)  Moulijn, J. A . ; van de Nouland, B. M . React. Kinet. Catal. Lett. 1975, 3, 405-408.  (23)  Giezynski, R.; Korda, A . J. Mol. Catal. 1980, 7, 349-354.  (24)  Wang, W.; Fujiki, M . ; Nomura, K. J. Am. Chem. Soc. 2005,127, 4582-4583.  (25)  L i , X . ; Baldamus, J.; Nishiura, M . ; Tardif, O.; Hou, Z. Angew. Chem. Int. Ed 2006, 45, 8184-8188 and references cited therein.  (26)  Longo, P.; Freda, C ; Ruiz de Ballesteros, O.; Grisi, F. Macromol Chem. Phys. 2001, 202, 409-412 and references cited therein.  (27)  Heiser, D. E.; Okuda, J.; Gambarotta, S.; Mülhaupt, R. Macromol Chem. Phys. 2005, 206, 195-202.  (28)  a) Natori, I. Imaizumi, K.; Yamagishi, H.; Kazunori, M . J. Polym. Sci. Part B: Polym. Phys. 1998, 36, 1657-1668. b) Imaizumi, K.; Natori, I. Nippon Kagaku Kaishi 1999, 5, 293-304 (abstract available in English).  Chapter 2. The Organic Products: Oligomers of Cyclohexene and 1,4-Cyclohexadiene^  A versión of this chapter has been submitted for pubHcation. Buschhaus, M . S. A . ; PampHn, C. B.; Blackmore, I. J.; Legzdins, P. Transformations of Cyclic Olefíns Mediated by Tungsten Nitrosyl Complexes. Reproduced in part with permission fi-om Organometallics, submitted for publication. Unpublished work copyright 2008 American Chemical Society.  2.1 Introduction In the past, exploration of the chemistry of nitrosyl-containing complexes in the Legzdins group has focused almost exclusively on the organometallic products produced. Recently, with the commencement of this project, the investigations of the reactivity of Cp'W(NO)(CH2CMe3)2 (Cp' = Cp* or Cp) with cyclic olefins, particularly with cyclohexene, have expanded our focus to include the organic products produced as well. This chapter presents the work done to characterize the ring-retaining oligomeric products derived from the cyclic olefins. Subsequent chapters will retum to considerations of the organometallic products and mechanistic insights into the reactivity of these systems.  2.2 Results and Discussion  2.2.1 Origin of the cyclohexene and 1,4-cyclohexadiene oligomers Ring-retaining oligomers derived from cyclohexene and 1,4-cyclohexadiene are produced when the tungsten precatalysts illustrated in Scheme 2.1 are thermolyzed in the neat cyclic olefins. The cyclohexene oligomers have been the most extensively characterized and are the primary focus of the following discussion. However, the 1,4-cyclohexadiene oligomers offer interesting insights and contrasts, which will be noted. As well, other substrates such as cyclopentene and cyclooctene are briefly examined.  Scheme 2.1  Thermolysis of Cp*W(NO)(CH2CMe3)2 (1) or CpW(NO)(CH2CMe3)2 (2) in neat cyclohexene at 70 °C for 40 h, or at 100 °C for 24 h, produces oligomers of cyclohexene. These  oligomeric products can be separated from the reaction mixture by chromatography on alumina with pentane or hexanes to yield a clear, colorless, moderately viscous oil. Subsequent investigations have revealed that Cp*W(NO)(CH2SiMe3)(ri^-CPhCH2) (5) and Cp*W(NO)[CH(Ph)CH2CH("Pr)CH2] (6), which is derived from 5, produced similar oligomers when thermolyzed at 100 °C for 24 h. In general, the composition and distribution of the cyclohexene oligomers in the mixture is similar regardless of the precatalyst used. However, the catalyst activities and the yields of oligomer show slight variations, as summarized in Table 2.1. As well, small amounts of the oligomer incorpórate an end group (CH2CMe3 or CH=CHPh) that is dependent on the precatalyst. Characterization of the oligomeric mixtures is based mainly on ' H N M R spectroscopy, low-resolution EIMS, and gas chromatography with mass spectrometry (GC/MS).  Table 2.1 Tumover frequencies for precatalysts 1, 5 and 6 at concentrations of 0.01 M in the oligomerization of cyclohexene. Precatalyst  Temperature (°C)  Turnover Frequency (/h) ^  1  70  1.7  1  100  6.2  5  100  5.5  6  100  6.5  transfer hydrogenation processes were not taken into account when the number of molecules of cyclohexene oligomerized was calculated.  2.2.2 Physical data for the cyclohexene and 1,4-cyclohexadiene oligomer mixtures  2.2.2.1 ' H N M R spectroscopy * H N M R analysis of the cyclohexene oligomer mixture gives a spectrum with complex peaks in the región of 1 - 2 ppm and in the región of 5.4 - 5.8 ppm. These regions correspond to the presence of aliphatic and olefínic protons, indicating that the oligomers retain unsaturated positions (double bonds). Integration of the two regions gives an average aliphatic to olefínic protón ratio of 12:1. By contrast, a linear oligomer obtained from a ring-opening metathesis  polymerization process gives an aliphatic to olefínic protón ratio of 4:1 regardless of chain length. Together with other evidence, particularly G C / M S analyses (vide infrd), the experimental ratio suggests that the ring-motif of the cyclohexene substrate is retained in the oligomers.  2.2.2.2 Low-resolution El-mass spectrometry Low-resolution EIMS analysis of the cyclohexene oligomer mixture yields a mass spectrum showing groups of peaks separated by 80 to 82 amu (Figure 2.1). In order to detect the higher masses, the temperature at which the spectrum is coUected must be ramped from 150 °C to 320 °C. The largest peak in each group of peaks is tabulated in Table 2.2 and compared to the degree of oligomerization and to the predicted mass for a mono-unsaturated ring-retaining cyclohexene oligomer. It is evident that as the oligomer size increases the level of unsaturation increases as well. The highest oligomer mass detected for a cyclohexene oligomer corresponds to a decamer with five unsaturations. (The term unsaturation, here and following, is used to refer only to the presence of double bonds in the oligomers.)  Table 2.2 Tabulated mass spectral data for cyclohexene and 1,4-cyclohexadiene oligomers. Cyclohexene oligomers  1,4-cyclohexadiene oligomers  n  Predicted mass  Observed mass^  Predicted mass  Observed mass''  2  164  163  160  160(161)  3  246  243  240  241(244)  4  328  325  320  321 (324)  5  410  405  400  402(409)  6  492  487  480  483 (487)  7  574  568  560  563 (568)  8  656  650  640  643 (649)  9  738  730  720  724 (728)  10  820  812  800  803 (N/A)  ^ largest peak in the group listed.  largest peak in the group usted with the highest miz peak listed in brackets.  Figure 2.1 Low-resolution El-mass spectrum of the cyclohexene oligomers (320 °C). Oligomer m/z = 163, 243, 325, 405, 487, 568, 650, 730 and 812.  s  a o o  (U  o  S8s^ig§gg^=  |g§&gg?ggs«  ggg¿g1t¿sé¿  Figure 2.2 Low-resolution El-mass spectrum of the 1,4-cyclohexadiene oligomers (150 °C). Oligomer m/z = 160, 241, 321, 402, 483, 563, 643, 724 and 803. Oligomer capped with CH=CHPh w/z = 181, 261, 344, 422, 505, 585, 667 and 746.  Interestingly, an EIMS analysis of the 1,4-cyclohexadiene oligomers (Figure 2.2) shows similar oligomerization levéis, with a decamer as the highest detected oligomer. This time, however, the oligomers have gained mass compared to predicted valúes (Table 2.2). Thus the second highest oligomer mass detected for a 1,4-cyclohexadiene oligomer corresponds to a nonamer with at least four fewer unsaturations than expected. When the oligomerization of 1,4-cyclohexadiene is performed with precatalyst 5, the mass spectrum of the organic product mixture shows peaks with masses that correspond to oligomers capped with a vinyl end group, CH=CHPh, in addition to the peaks representing oligomers derived only from 1,4-cyclohexadiene (Figure 2.2). The corresponding neopentylcapped oligomers (neopentyl = CH2CMe3) from precatalyst 1 do not give signáis in the EIMS spectrum, but they do give rise to peaks in the GC/MS spectrum {vide infra). Such capped oligomers, varying with the precatalyst employed, derive from the initiation of the catalyst. Further evidence from the related molybdenum systems supports the proposed initiation mechanism (yide infra, Chapters 3 and 5).  2.2.3 Transfer dehydrogenation in the oligomerization systems The varying levéis of unsaturation in the cyclohexene and 1,4-cyclohexadiene oligomers (tabulated in Table 2.2) are due to a transfer dehydrogenation process, as demonstrated by ' H N M R spectroscopic experiments performed by Dr. lan Blackmore. When the in situ reaction mixture derived from the thermolysis of precatalyst 5 in cyclohexene at 60 °C for 24 h is examined by ' H N M R spectroscopy, a peak due to cyclohexane appears at 1.83 ppm in the spectrum, and higher-than-predicted levéis of unsaturation appear in the cyclohexene oligomers {vide supra). Figure 2.3 illustrates the ' H N M R spectrum of this reaction mixture. Conversely, similar analysis of the reaction mixture of 5 in 1,4-cyclohexadiene shows the appearance of a peak at 7.16 ppm due to the formation of benzene, and a small peak at 1.37 ppm due to cyclohexane, as illustrated in Figure 2.4. The 1,4-cyclohexadiene oligomers contain lower levéis of unsaturation than predicted.  Figure 2.3 ' H N M R spectrum of the reaction mixture of 5 in neat cyclohexene after 24 h (300 MHz, cyclohexene, rt) demonstrating the concomitant formation of cyclohexane with the cyclohexene oligomers.  2  Figure 2.4 ' H N M R spectrum of the reaction mixture of 5 in neat 1,4-cyclohexadiene after 24 h (300 MHz, 1,4-cyclohexadiene, rt) demonstrating the concomitant formation of benzene with the 1,4-cyclohexadiene oligomers.  The net results of the transfer dehydrogenation process are illustrated in Scheme 2.2. The catalytically active tungsten species derived from the precatalysts and capable of oligomerization may also be responsible for the transfer dehydrogenation processes. In the case of the 1,4cyclohexadiene oligomerization, the formation of aromatically-stable benzene likely provides the driving forcé for the transfer of hydrogen from 1,4-cyclohexadiene to the oligomers. The driving forcé for the reverse transfer dehydrogenation reaction when cyclohexene is the substrate is not as clear. It may be the dehydrogenation of the oligomers or the hydrogenation of cyclohexene to cyclohexane.  Scheme 2.2  2.2.4 GC/MS analyses of the cyclohexene and 1,4-cyclohexadiene oligomer mixtures  2.2.4.1 GC data for the cyclohexene oligomer mixture Analysis of the cyclohexene oligomers by gas chromatography with a mass spectrometer detector (GC/MS) gives further Information about the composition of the oligomer mixture. The G C analyses use a non-polar, cross-linked 5% diphenyl- 95% dimethylpolysiloxane column and a standard-method temperature gradient that rises from 120 °C to 300 °C. A typical G C trace of the cyclohexene oligomers derived from precatalyst 1 has clusters of peaks, as shown in Figure 2.5. Based on the parent masses of the peaks, the clusters are identifíed with discrete oligomer lengths (families). Cyclohexene dimers are represented at 1.5 min, trimers at 9-11 min, tetramers at 13-14, pentamers at 15.5-17.5 and higher oligomers any where from 19-24 min. Oligomers higher than pentamers are difficult to detect with GC  techniques because they do not come through the column easily and they make up only a small fraction of the total oligomer volume. In addition to the peaks due to the oligomer families, peaks due to dimers and trimers capped with a neopentyl group emerge at 6.2 min and at about 12 min, respectively. Each peak within a family cluster represents a unique compound. Thus, the more than eleven peaks in the trimer cluster (Figure 2.5) correspond to more than eleven different trimeric oligomers. The differences are in ring coimectivity and in unsaturation levéis.  Jt)unámvi6  4000000  300000  3000000  2600000  2000000  1500000  1000000  BOOWK)  V Tfm8->  2.00  400  6.00  8.00  10.00  12.00  14.00  16.00  18.00  20.00  22.00  24v00  Figure 2.5 A typical G C trace for the mixture of cyclohexene oligomers (min, standard method).  Integration of the G C peaks gives Information about the relative percentages of each type of cyclohexene oligomer present in the mixture derived from precatalyst 1. Average valúes from the best analyses indícate a distribution of about 7% dimers, < 1 % dimer capped with neopentyl, 45% trimers, negligible trimer capped with neopentyl, 36% tetramers, 10% pentamers, and < 1% higher oligomers. In cyclohexene oligomer mixtures derived from precatalyst 5, the GC trace looks similar to that of mixtures derived from precatalyst 1, and the integrated percentages for each oligomer family remain similar. However, none of the other peaks in the GC trace can be defmitively identified as arising from oligomers capped with a CH=CHPh end group.  2.2.4.2 MS data for the cyclohexene oligomer mixture The mass spectra associated with each peak in the GC trace all tell a similar story. The individual cyclohexene oligomers fragment in a distinctive pattem that indicates the loss of 6carbon units, as illustrated in Figures 2.6 through 2.9 for each of the oligomer lengths. This fragmentation pattem suggests that the ring stmcture of the monomer is retained in the oligomer. If the rings had been opened by ROMP, the expected fragmentation pattem would be that of a linear unsaturated hydrocarbon, namely loss of C„H2„ units with an relatively large proportion of 2-carbon and 3-carbon units lost.' In addition, Figure 2.10 shows the fragmentation pattem of a cyclohexene dimer capped with a neopentyl group. This compound also fragments in a manner consistent with retention of the ring stmcture, losing either a cyclohexyl or a neopentyl group to form species that give rise to the peaks at 151 and 163, respectively.  Abundtnot  Vi 200000 164 ISOOOO  87 se  10ÚO0O  HOOO  so  80  Ido  lao  140  lio  leo  ,.1^, a30  2 4 0 Z U 3 8 0  300  ,3^.yf. 3 Z 0 3 4 0 3 S 0  380  400  420  Figure 2.6 A typical G C mass spectrum of a cyclohexene dimer (MS m/z = 164 [P"^], GC peak at t= 1.4 min).  2U  tS2  149  1JS  fTVX->  Figure 2.7 A typical G C mass spectrum of a cyclohexene trimer (MS m/z = 246 [P^], G C peak at t = 9.7 min).  HOOOO  lío 3 »  U1  100000  •I  m  m  m  %m m  m  m  m  tm tm V» im -m m  vfí  OS)  Figure 2.8 A typical G C mass spectrum of a cyclohexene tetramer (MS m/z = 326 [P"^], GC peak att= 13.9 min).  IB3  243  m  4oe  3»  ios JM  1*0  1 »  1»0  JOO 220  í*0  700  sao  300  MO  140  3»436<  3M384  3«  400  no  Figure 2 9 A typical G C mass spectrum of a cyclohexene pentamer (MS m/z = 408 [P*], G C peak at t = 16.1 min).  Mwndinc» 80000  60000 57  180  SI  234 «7  «  60  4  eo  107 100  ' 2 ' I3S 120  I  T188  203 218  244ZM 2M27e2te  3K  40Z  t 4 0 l W 1 8 0 ^ M r M Q ^ ^  Figure 2.10 A typical G C mass spectrum of a cyclohexene dimer capped with a neopentyl end group (MS m/z = 234 [P^], GC peak at t = 6.2 min).  The mass spectra obtained during the G C analyses also show that the unsaturation levéis vary within each oligomer group. Thus, for the dimer cluster, three distinct parent masses of 166, 164 and 162 are identifíed that correspond to zero, one and two unsaturations per dimer. Similarly, in the trimer cluster each peak's parent mass corresponds to either one or two unsaturations present in each unique trimer. Table 2.3 correlates each cyclohexene oligomer family with the number of peaks in its GC spectrum cluster, its G C peak times, the variable M S  parent masses, and the empirical formula with the number of unsaturations that this represents for the cyclohexene oligomer mixture derived from precatalyst 1.  Table 2.3 Summary of GC and M S data for the cyclohexene oligomer families. Parent masses  Empirical formulae  Number of unsaturations  1.35-1.49  166, 164, 162  Ci2H22,20,18  0,1,2  1  6.22  234  C17H30  1  Trimers  11-16  9.25-9.98  246, 244  Ci8H3o,28  1,2  Tetramers  15-16  13.07-13.96  326,324  C24H3g36  2,3  Pentamers  4-12"  15.80-17.69  408, 406  C3oH48,46  2,3  Higher  3-12''  19.63-22.60  488+'  C36H56  3 or more  Oligomer group  Number of peaks  Range of peak times (min)  Dimers  2-3  Dimer + npt*  ^ npt = neopentyl end group. the number of peaks visible above the baseline noise varies greatlyfromsample to sample. ° the MS detects from O to 500 amu, so the exact parent masses beyond 488 are unknown.  2.2.4.3 GC and MS data for the 1,4-cyclohexadiene oligomer mixture GC/MS analysis of the 1,4-cyclohexadiene oligomer mixture reveáis clusters of broad peaks that can be attributed to dimers (miz = 162, 160), trimers {miz = 244, 242), tetramers {miz == 322, 320), pentamers {miz = 400), and higher oligomers {m/z = 480). The clusters appear at times of 1.9-3.0 min, 10.1-11.2 min, 13.4-14.8 min, 15.9-17.9 min, and 18.7-24 min, respectively, and integration of the peaks gives relative percentages of 3% dimers, 18% trimers, 37% tetramers, 29% pentamers and 12% higher oligomers. The fragmentation pattems in the M S spectra show the loss of 79-83 amu groups, which indicates the loss of 6-carbon units with variable unsaturation levéis, typically one or two double bonds. Compared to the analogous cyclohexene oligomer mixture, the 1,4-cyclohexadiene oligomer mixture contains a higher percentage of the heavier oligomers, such as tetramers and pentamers.  2.2,5 Concentration effects in the C p * W ( N O ) ( C H 2 C M e 3 ) 2 system The amount of oligomer produced per mole of precatalyst 1 varies with the initial concentration of 1 in cyclohexene. Thus, under the typical reaction conditions of 70 °C for 40 h, as the concentration of the solution decreases the relative yield of oligomer per mole of tungsten  catalyst increases. For example, at a concentration of 0.054 M , 21 moles of cyclohexene are converted per mole of tungsten. As the concentration decreases to 0.026 M , 0.010 M , 0.0059 M and 0.0036 M , the moles of cyclohexene converted per mole of tungsten increase to 37, 69, 83 and 82, respectively, as summarized briefly in Table 2.4 (and more fully in Table 2.5). These tumover numbers show that the system is catalytic, not stoichiometric, in its conversión of cyclohexene to oligomer. Although the catalyst loadings are in the low range desirable for catalysts generally, the tumover frequencies (0.5, 0.9, 1.7, 2.1, and 2.0 mol/h for these samples) are unremarkable. The inverse relationship between concentration and cyclohexene conversión suggests that the lower concentrations discourage a catalyst decomposition pathway involving múltiple tungsten centers.  Table 2.4 Concentration effects on the moles of cyclohexene converted per mole of tungsten. Concentration (M)  0.054  0.026  0.010  0.0059  0.0036  0.108  0.104  0.120  0.029  0.018  Volume of cyclohexene (mL)  4.1  8.2  24  10  10  Mass of isolated oligomer (g)^  0.379  0.634  1.383  0.406  0.239  21  37  69  83  82  11.4  9.5  7.1  5.0  3.0  0.54  0.26  0.10  0.06  0.04  Mass of Cp*W(NO)(CH2CMe3)2 (g)  Moles of cyclohexene converted per mole of tungsten Amount of cyclohexene converted to oligomer (wt %) Catalyst loading (mol %) After thermolysis at 70 °C for 40 h for all samples.  The thermolysis of 1 in cyclohexene has also been performed at lower reaction temperatures and for longer reaction times, as summarized in the latter part of Table 2.5. In terms of the moles of cyclohexene converted to oligomer per mole of tungsten catalyst, an increase in reaction time seems to have no effect, and a decrease in reaction temperature has a negative impact. In contrast, thermolysis of 1 in cyclohexene at 100 °C for 24 h increases the oligomer yield, and thus the tumover number rises to 149 and tumover frequency to 6.2 mol/h.  Table 2.5 The effect of initial concentration on the moles of cyclohexene converted per mole of tungsten precatalyst 1 at various reaction temperatures and reaction times. Concentration (M)  0.0536  0.0258 0.0102 0.0083 0.0059  0.0036  0.0010  0.0001  Time (h)  40  40  40  40  40  40  40  40  Temperature (°C)  70  70  70  70  70  70  70  70  0.108  0.104  0.120  0.041  0.029  0.018  0.005  0.001  4.1  8.2  24  10  10  10  9  10  0.501  0.751  1.516  0.585  0.440  0.260  0.072  0.006  0.379  0.634  1.383  0.564  0.406  0.239  0.051  0.004  0.0046  0.0077  0.0168  0.0069  0.0049  0.0029  0.0006  0.00005  21.0  36.5  68.9  82.5  83.2  81.6  67.8  47.9  0.544  0.262  0.103  0.084  0.060  0.036  0.010  0.001  Mass of 1 (g) Volume of cyclohexene (mL) Mass of crude products (g) Mass of isolated oligomer (g) Moles cyclohexene in oligomer (mol) Moles cyclohexene converted per mole of tungsten catalyst Catalyst loading (mol %) Table 2.5 con't.  Concentration (M)  0.0108 0.0091 0.0080 0.0200 0.0160 0.0100 0.0040  0.0103  Time (h)  65  115  115  115  115  115  115  24  Temperature (°C)  70  70  70  55  55  55  55  100  0.053  0.049  0.039  0.098  0.079  0.049  0.020  0.101  10  11  10  10  10  10  10  20  0.778  0.564  0.494  0.515  0.487  0.316  0.096  2.655  0.745  0.503  0.466  0.408  0.410  0.267  0.081  2.523  0.0091  0.0061  0.0057  0.0050  0.0050  0.0033  0.0010  0.0307  84.1  61.2  70.8  24.8  31.0  32.5  24.6  149.4  0.109  0.092  0.081  0.203  0.163  0.102  0.041  0.104  Mass of 1 (g) Volume of cyclohexene (mL) Mass of crude products (g) Mass of isolated oligomer (g) Moles cyclohexene in oligomer (mol) Moles cyclohexene converted per mole of tungsten catalyst Catalyst loading (mol %)  2.2.6 Bulk separations of the cyclohexene oligomers  2.2.6A Unsuccessful attempts in the separation of the cyclohexene oligomers Various methods have been employed in the attempts to sepárate the mixture of cyclohexene oligomers into its component parts, whether into the oligomer families or into individual compounds. Most of these methods do not give the desired results. Selective solvation experiments with methanol and acetontrile provide evidence of a partial separation between the components that dissolve in the solvent and those that do not. Thus, as determined by G C / M S analysis, the cyclohexene oligomers recovered from the solvents were somewhat enriched in the lower oligomers (dimers, trimers) while the oligomers that did not dissolve were enriched in the higher oligomers, as tabulated in Figure 2.11. However, the levéis of enrichment obtained by this method are not sufficient to give samples suitable for further analysis.  100% 1 I Pentamers and higher oligomers I Tetramers  n Trimers  O Dimers  Sample Numtier  Figure 2.11 Relative oligomer percentages obtained fi-om selective solvation experiments. Sample 1 - original cyclohexene oligomers; Sample 2 - cyclohexene oligomers soluble in acetonitrile; Sample 3 - cyclohexene oligomers insoluble in acetonitrile; Sample 4 cyclohexene oligomers soluble in methanol; Sample 5 - cyclohexene oligomers insoluble in methanol.  Analytical and preparative high-pressure liquid chromatography ( H P L C ) experiments have given no evidence that the cyclohexene oligomers can be separated by this method. The ' H  N M R spectra of fractions collected during the H P L C experiments do not match the ' H N M R spectrum of the original oligomer mixture. Therefore, it is doubtfiíl that any oligomers were in the collected fractions at all. When thin-layer chromatography (TLC) is attempted using either alumina or silica plates, with pentane as the mobile phase, the cyclohexene oligomers travel at the solvent front, without any separation whatsoever. T L C silica plates doped with AgNOa also give no significant separation of the oligomers, this time because the oligomers do not move on the píate.  2.2.6.2 Vacuum distillation of the cyclohexene oligomer mixture Vacuum distillation of the cyclohexene oligomer mixture gives reasonable separation of the lowest oligomer families. At 10'^ mm Hg, two fractions distill at 55 °C and 116-120 °C, respectively. G C / M S analyses of these fractions and of the undistilled oligomers allow Identification and quantifícation of their components, as illustrated in Figure 2.12. Thus, the fírst fraction to distill contains 100% cyclohexene dimers. The second fraction contains up to 94% trimers, with some dimer and tetramer contamination. The oligomers that remain undistilled are a combination of trimers (20%), tetramers (56%)), and higher oligomers (24%)).  100  «  80  í o E o  60  5 O o O)  • Dimer  -  40  _ • Tetramer  20  BRentamer  ti í  H Trimer  Distillation fraction  Figure 2.12 Composition of the vacuum-distillation fractions of the cyclohexene oligomers, with 1 = fírst fraction (distills at 55 °C), 2 = second fraction (distills at 116-120 °C), and 3 = still-pot contents (undistilled).  2.2.7 Further characterization of the cyclohexene dimers and identifícation of the major mono-unsaturated cyclohexene dimer The cyclohexene dimers isolated by vacuum distillation have been further characterized by N M R analyses and by hydrogenation experiments. The ' H N M R spectrum of the distilled cyclohexene dimers (Figure 2.13) shows complex aliphatic peaks at 0.8 - 2.1 ppm and olefmic peaks at 5.58 - 5.75 ppm. Integration of these regions gives a ratio of 7:1 for the aliphatic to olefmic protons. Theoretically, a ring-retaining dimer with one unsaturation would give a ratio of 9:1, and a dimer with two unsaturations would give a ratio of 3.5:1. The experimental valué falls between these two theoretical valúes, suggesting an averaging over the entire dimer mixture which contains both species. This is supported by the GC/MS data which contain a small peak with a parent mass for a dimer with no unsaturations and larger peaks with parent masses for dimers with one and two unsaturations. The '^C N M R spectrum for the cyclohexene dimer mixture (Figure 2.14) shows at least 13 olefínic carbón peaks between 127 ppm and 132 ppm. This feature suggests the presence of several unsaturated species in the mixture, and likely reflects the variety of isomers as well as the differing unsaturation levéis.  pp«  Figure 2.13 ' H N M R spectrum of the distilled cyclohexene dimer mixture (600 MHz,  CóDó,  rt).  Figure 2.14 ~C N M R spectrum of the distilled cyclohexene dimer mixture (600 MHz, CóDe, rt).  Hydrogenation of the cyclohexene dimer mixture using Wilkinson's catalyst yields the fully saturated dimer, and subsequent N M R analyses confirm the ring-retaining structure of the oligomer. The ' H N M R spectrum. Figure 2.15, shows the disappearance of the olefínic-proton peaks. The remaining aliphatic-proton peaks are complex multiplets, the specific identities of which can be assigned with the aid of a ' H - ' H C O S Y N M R experiment. This assignment is consistent with retention of the cyclohexyl ring, and thus the saturated dimer is identified as cyclohexylcyclohexane. The '•'C N M R spectrum (Figure 2.16) is also consistent with this identification, with three peaks appearing in the spectrum. The peak at 43.8 ppm corresponds to the methine carbons of the coupled rings. The peak at 30.5 ppm is due to the methylene carbons next to the methine carbons, and the peak at 27.3 ppm represents the other three methylene carbons of each ring.  i re 2.5 ' h N M R s p e c , ™ of eyc,„hex„cyclohexane produced ly a,e c o ^ p l e . hydrogenation of the cyclohexene dimer mixture (300 MHz, Q D , , rt).  «a  Figure 2.16  •20  •  •  T ,¿0  ^  íd  N M R spectrum of cyclohexylcyclohexane produced by the complete  hydrogenation of the cyclohexene dimer mixture (300 M H z , C , D , , rt).  Further information has been obtained from a second, incomplete hydrogenation reaction that was run for only 3 h. The ' H and '^C N M R  spectra, shown in Figures 2.17 and 2.18, reveal  the presence of two products, present in a 1:2 ratio. The first of these is identified as cyclohexylcyclohexane, on the basis of comparison of the ' H and '•'C N M R data with the spectra obtained in the complete hydrogenation experiment. The second product is an unsaturated cyclohexene dimer, as shown by the olefínic peaks which appear in the N M R Specifically, the ' H N M R ppm,  spectra.  spectrum displays two complex multiplets between 5.63 and 5.74  each of which integrates for 1 H . The splitting pattems of these multiplets are complicated  by the A B pairing of the peaks, present even at a 600-MHz field strength, but clearly they are different from each other, roughly resembling a broad doublet and a doublet of doublets (Figure 2.19a). The '^C N M R  spectmm displays two olefínic peaks at 127.4 ppm and 131.2 ppm (Figure  2.19b). Comparison of these peaks to those appearing in the N M R  spectra of the full  cyclohexene dimer mixture reveáis that this second compound is the major isomer of the monounsaturated dimer.  BOU  r-  /  Figure 2.17 H N M R  -r- ,  1 f>  ^  8  T^.-  4  ^ , - ~ - - - , — ^ - r - ' - — r  r ?  r  -f1  spectrum of cyclohexylcyclohexane and 3-cycIohexylcyclohexene  produced by the partial hydrogenation of the cyclohexene dimer mixture (600 MHz, C^De,  rt).  jJuJL -*T—  —  —r-  100  Figure 2.18  40  N M R spectrum of cyclohexylcyclohexane and 3-cyclohexylcyclohexene  produced by the partial hydrogenation of the cyclohexene dimer mixture (600 MHz, CeDf,, rt).  (a)  »J0  S.»  S.TO  MS  5«>  S.»  S.«  (b) '  in  >»  at  ue  Figure 2.19 Expansions showing the characteristic olefmic peaks of 3-cyclohexylcyclohexene in (a) the ' H N M R spectrum and (b) the '^C N M R spectrum (600 M H z , CóDe, rt).  The unsaturated cyclohexene dimer is identifíed as 3-cyclohexylcyclohexene, based on the N M R spectroscopic data. Integration of the ' H N M R peaks suggests that the dimer is monounsaturated. There are thus four possible candidates, illustrated as options A , B, C and D in Scheme 2.3. The presence of two inequivalent olefmic peaks in the '^C N M R spectrum rules out  the symmetrical option A . The presence of two olefmic protons, as evidenced by the H N M R spectrum, rules out option B . The differing splitting pattems of the olefmic protón peaks suggest that option C is a better candidate than option D. The presence of the cyclohexyl substituent at the 3-position in option C leaves only one aliphatic protón to interact with the nearest olefmic protón instead of two, thus changing the splitting pattem of that protón. With the help of ' H - ' H COSY, 'H-'^C H M Q C  and ' H - ' ^ C H M B C N M R experiments, the entirety of the ' H and ' ^ C  N M R spectra can be consistently assigned to 3-cyclohexylcyclohexene. Thus the major isomer of the mono-unsaturated cyclohexene dimers is conclusively identified.  Scheme 2.3  2.2.8 Other cyclic olefín substrates The ability of the tungsten precatalysts to oligomerize cyclic olefins extends to other substrates. Thus, cyclopentene has been oligomerized with precatalysts 1, 2 and 5 (by Dr. Craig Pamplin, myself, and Dr. lan Blackmore). In the case of precatalysts 1 and 5, EIMS analyses show the formation of oligomers up to and including dodecamers {miz = 810). Like the oligomers of cyclohexene, the oligomers of cyclopentene also show increased levéis of unsaturation (up to four double bonds in the longest oligomers). Cyclooctene is oligomerized in small amounts by precatalyst 1. A low-resolution EIMS of the oligomeric products reveáis the presence of dimers through septamers of cyclooctene. Precatalyst 1 also oligomerizes 4-methylcyclohexene. G C / M S analysis on the isolated 4methylcyclohexene oligomers gives clear evidence for oligomers up to and including tetramers. Thus, the presence of the methyl substituent does not stop the oligomerization, but it does influence the length and distribution of the oligomers. 1,3-cyclohexadiene does not oligomerize under thermolysis conditions with precatalyst 5. This may be due to the ability of the substrate to form a complex with the tungsten center, in  analogy to previous work that has shown the Cp*W(NO) fragment forms very stable complexes when exposed to 1,3-dienes such as 1,3-butadiene? On the other hand, this negative result with 1,3-cyclohexadiene indicates that 1,4-cyclohexadiene is not isomerized.  2.2.9 Comparisons of the cyclohexene oligomerization products obtained by thermolysis of 1 and 5 with the products produced by other systems The cyclohexene-derived products of precatalysts 1 and 5 can be compared with the products obtained from the traditional metathesis systems discussed in Chapter 1. Like those past systems, no evidence for R O M P is observed and the cyclohexene oligomers retain the ring motif in their structure. However, there are key differences between the systems. The Re system^ offers little correlation with our system since the products are quite different, namely saturated cyclohexene polymers with molecular weights of 2500 amu versus unsaturated cyclohexene oligomers with molecular weights of 164-812 amu. The cyclohexene ring structure is retained in both systems. However, the Re system empirically has a bondopening result (to give full saturation), which is quite different from the mechanisms believed to be operative in the Legzdins tungsten system (see Chapter 5). The 1970-80 systems based on tungstenhexachloride yield oligomeric cyclohexene products of similar size range to our system. However, the levéis of saturation differ. In the WCl6 + Me4Sn system, ' H N M R analysis of the cyclohexene oligomers reports integration ratios of aliphatic to olefínic protons of 27:1 for dimers, 45:1 for trimers and 170:1 for the higher oligomers.'* In contrast, our system yields a ratio of 6.5:1 for the isolated dimers and 12:1 for the overall oligomer mixture. Clearly, our tungsten nitrosyl system creates cyclohexene oligomers with much higher levéis of unsaturation than those previously reported. The WCle + ROH + EtAlCb (R = Et, Ph, PhCHi) system^ reports cyclohexene oligomers that are mainly saturated, and identifíes the most abundant dimer as the saturated cyclohexylcyclohexane. Although a monounsaturated dimer is claimed, the exact identity is not determined. The complete lack of mechanistic insight into the early systems is a serious drawback. The catalytic precursurs could not be detected, mush less the active species in the reaction soups. In contrast, our system allows insight into the initiation pathways and the proposal of a feasible catalytic cycle consistent with the observed products of our reactions.  2.3 Summary Ring-retaining oligomers of various cyclic olefíns are formed as the organic products in thermolysis reactions of a selection of tungsten precatalysts (1, 2, 5 and 6) in neat cyclic olefins. Cyclohexene and 1,4-cyclohexadiene form oligomers up to and including decamers, with a predominance of trimers and tetramers, or tetramers and pentamers, respectively. A transfer dehydrogenation process increases or decreases the expected levéis of unsaturation in the oligomers, dependent on the cyclic olefin substrate utilized. The cyclohexene oligomers have higher levéis of unsaturation than any other similar mixture reported in the literature, and thus offer a greater potential for further functionalization. However, it remains to be seen i f industrial interest in such functionalized oligomers will be strong enough to justify further research in this área. When the oligomerizations are performed with precatalysts 1 or 2, the well-known alkylidene intermediates do not interact with the cyclic olefin in a ring-opening metathesis polymerization (ROMP) type mechanism. Instead, the ring-motif is retained in the resulting oligomers, as seen in the G C / M S analyses and in the Identification of the major monounsaturated dimer isomer as 3-cyclohexylcyclohexene. Small amounts of neopentyl-capped cyclohexene oligomers are also formed, hinting at an initiation process that couples the neopentyl ligand from the tungsten precatalyst to a molecule of the cyclic olefin. The mechanism by which the ring-retaining cyclic-olefin oligomers are formed will be considered in Chapter 5. The proposed mechanism must be able to account for the di verse observations detailed in this chapter and also for the organometallic products of the molybdenum and tungsten nitrosyl complexes respectively described in the next two chapters.  2.4 Experimental Procedures  2.4.1 General methods A l l reactions and subsequent manipulations involving organometallic reagents were performed under anaerobic and anhydrous conditions either at a vacuum-nitrogen dual manifold or in an inert-atmosphere dry box. Pentane, hexanes, benzene-í/^, diethyl ether, and tetrahydrofuran (THF) were all dried over sodium benzophenone ketyl and were freshly distilled prior to use. Cyclohexene and 1,4-cyclohexadiene were purchased from Aldrich and Acros Organics, respectively, dried over sodium benzophenone ketyl, distilled, and stored in resealable glass vessels. Cp*W(NO)(CH2CMe3)2 (1),^ Cp*W(NO)(CH2SiMe3)(Ti^-CPhCH2) (5)^ and Cp*W(NO)(CHPhCH2CH("Pr)CH2) (6)^ were prepared according to the published procedures. CpW(NO)(CH2CMe3)2 (2) was not prepared according to the original method^" but according to a modified versión of the Cp*Mo(NO)(CH2CMe3)2 preparation.^" A l l other chemicals were purchased from Aldrich and used as received. N M R spectra were recorded at room temperature on Bruker AV-300, Bruker AV-400 or Bruker AV-600 M H z Instruments using standard U B C pulse sequences with delay times of 1 sec for ID experiments and 1.5 sec for 2D experiments. A l l chemical shiñs and coupling constants are reported in ppm and Hz, respectively. *H N M R spectra were referenced to the residual protio isotopomer present in CQDS (7.15 ppm). '^c N M R spectra were referenced to CeDe (128 ppm). Where necessary, ' H - * H C O S Y , ' H - ' ^ C H M Q C , and ' H - " c H M B C experiments were carried out to correlate and assign ' H and '''C N M R signáis. G C / M S analyses were carried out on an Agilent 6890 Series G C system equipped with a non-polar, cross-linked 5% diphenyl- 95% dimethylpolysiloxane column and an Agilent 5973 Network mass-selective detector. Lowresolution mass spectra (El, 70 eV) were recorded by the staff of the U B C mass spectrometry facility using a Kratos MS-50 spectrometer.  2.4.2 Oligomerization of cyclohexene and 1,4-cyclohexadiene with 1, 2, 5 and 6 Thermolyses of 1 and 2 in neat cyclohexene produced oligomers of cyclohexene, which were generally isolated and analyzed according to the following method, the procedure for cyclohexene with 1 being described as a representative example. Compound 1 (0.108 g) was dissolved in cyclohexene (4.1 mL) to give a red solution (0.05 M). Thermolysis of this solution at 70 °C for 40 h yielded a dark brown solution. The vacuum-transferred volátiles were clear and  colorless. The reaction residue was a dark brown oil that flowed easily (0.501 g). The residue completely dissolved in pentane and was chromatographed on an alumina column (5 x 0.5 cm). The pentane fraction (ca. 10 mL) yielded a clear oil (0.379 g). EtaO and THF fractions were also collected, and they yielded dark brown residues (0.061 g and 0.021 g, respectively). The final column color was light orange-brown. GC/MS analyses and ' H N M R spectroscopy were performed on the isolated oil. The thermolysis of 1 in cyclohexene was carried out at various concentrations, reaction times and reaction temperatures. Representative data and results are summarized in Table 2.5. A blank of neat cyclohexene (4.0 mL) was thermolyzed at 70 °C for 40 h. The solution remained clear and colorless. A GC/MS analysis was performed. The oligomerizations of cyclohexene and 1,4-cyclohexadiene were effected with 5 in a similar manner by Dr. lan Blackmore. The procedure for cyclohexene with 5 is described as a representative example. A 0.010 M solution was prepared by dissolving 5 (0.027 g) in cyclohexene (5.0 mL). This solution was heated at 100 °C for 24 h during which time the solution changed color from red to brown. The volatile components were removed from the final mixture in vacuo to obtain a viscous brown oil that was re-dissolved in a mínimum of pentane and was transferred to the top of a silica column (ca. 7 x 0.5 cm). The organic products were then eluted from the column with a copious amount of pentane, and solvent removal from the eluate yielded a colorless oil (0.54 g). Similarly, Cp*W(NO)[CH(Ph)CH2CH("Pr)CH2] (6) (0.026 g in 5.0 mL cyclohexene) gave 0.64 g of colorless oil at 100 °C.  2.4.3 Physical data for the bulk cyclohexene and 1,4-cyclohexadiene oligomer mixtures NMR analysis of the cyclohexene oligomer mixture. ' H N M R (CóDe, 300 MHz, 25 °C) 5 0.7-1.3, 1.3-1.8, 1.8-2.2, 5.4-5.5, 5.5-5.8. These were all broad peaks made up of many complex overlapping signáis. No assignments beyond Identification of the regions as representing aliphatic or olefmic protons could be made. Integration over the regions of 0.7 - 2.2 ppm and of 5.4 - 5.8 ppm gave an average aliphatic to olefmic protón ratio of 12:1. EIMS analysis of the cyclohexene and 1,4-cyclohexadiene oligomer mixtures. Lowresolution EIMS spectra were recorded for the of cyclohexene oligomers obtained from the thermolysis of 1 in cyclohexene at 70 °C, and for the mixture of 1,4-cyclohexadiene oligomers  obtained from the thermolysis of 5 in cyclohexadiene at 50 °C. Note the envelopes of peaks corresponding to oligomers containing a vinyl end group (CH=CHPh) for the 1,4 cyclohexadiene oligomers. Cyclohexene oligomer mixture, largest peak in each cluster of peaks recorded: M S (El, 320 °C) miz 163, 243, 325, 405, 487, 568, 650, 730, 812  of oligomers.  Cyclohexadiene oligomer mixture, largest peak in each cluster of peaks recorded: M S (El, 320 °C) miz 160, 241, 321, 402, 483, 563, 643, 724, 803 [?""] of oligomers. miz 181, 261, 344, 422, 505, 585, 667, 746 [P^] of oligomers capped with CH=CHPh vinyl end group.  2.4.4 GC/MS analyses of the cyclohexene and 1,4-cyclohexadiene oligomer mixtures GC/MS samples were prepared by dissolving a few drops of the oligomer mixtures to be analyzed in EtaO. A standard method was used that began with 5 min at 120 °C, foUowed by a temperature ramp of 15 °C/min over 12 min up to 300 °C, and finally 15 min at 300 °C, for a total run time of 32 min. A G C / M S analysis of the cyclohexene blank showed only signáis due to cyclohexene, as expected. A G C / M S analysis of the volátiles removed from the cyclohexene oligomerization reaction with 1 showed a trace of the dimers among the cyclohexene. (A low resolution EIMS spectrum also confirmed the presence of a small amount of cyclohexene dimer, miz = 164.) Repeated G C / M S analyses of the cyclohexene oligomer mixtures obtained with precatalyst 1 consistently showed cyclohexene oligomers up to and including pentamers. Peaks due to dimers appeared at 1.5 min, and integrated in the range of 5.9-7.5% with an average of 7%. Dimer M S miz valúes were 166, 164 and 162. Peaks due to trimers appeared at 9-11 min, and integrated in the range of 38.7-49.1% with an average of 45%). Trimer M S miz valúes were 246 and 244. Peaks due to tetramers appeared at 13-14 min, and integrated in the range of 31.737.4% with an average of 36%. Tetramer M S miz valúes were 326 and 324. Peaks due to pentamers appeared at 15.5-17.5 min, and integrated in the range of 5.3-20.1% with an average of 10%. Pentamer M S miz valúes were 408 and 406. Higher oligomers (< 1%) were difficult to detect with this instrumentation. A dimer capped with a neopentyl unit gave rise to a peak at 6.2 min (< 1%), and the neopentyl-capped trimers gave rise to peaks at about 12 min. The relative percentages of the smaller peaks, such as pentamers and higher oligomers, varied greatly depending on the baseline noise and the concentration of the injected sample.  GC/MS analysis of the oligomers of cyclohexene obtained from a reaction of cyclohexene with 5 gave peak integration valúes of 7% dimers, 46% trimers, 34%o tetramers, 11%) pentamers, and 2% higher oligomers. The GC trace for the 1,4-cyclohexadiene oligomers showed broad clusters of overlapping peaks, which corresponded to oligomer families. The peaks for the dimers appeared at 1.9-3.0 min (3%)), the trimer peaks appeared at 10.1-11.2 min (18%)), the tetramer peaks appeared at 13.4-14.8 min (37%), the pentamer peaks appeared at 15.9-17.9 min (29%o), and higher oligomer peaks appear at 18.7-24 min (12%).  2.4.5 Monitoring of transfer dehydrogenation by *H NMR spectroscopy ' H NMR spectroscopic monitoring of the oligomerization of cyclohexene by 5 at 60 °C. A sample of 5 (10 mg) was dissolved in cyclohexene and heated at 60 °C for 24 h in a sealed N M R tube and monitored by ' H N M R spectroscopy. Figure 2.17 displays the spectrum after 24 h, clearly showing the presence of cyclohexane (1.83 ppm) and a trace of benzene (7.62 ppm). NMR spectroscopic monitoring of the oligomerization of 1,4-cyclohexadiene by 5 at 60 °C. A sample of 5 (10 mg) was dissolved in 1,4-cyclohexadiene and heated at 60 °C for 24 hours in a sealed N M R tube and monitored by ' H N M R spectroscopy. Figure 2.18 displays the spectrum after 24 hours, clearly showing the presence of benzene (7.16 ppm) and a trace of cyclohexene (1.37 ppm).  2.4.6 Bulk separation: Unsuccessful attempts in separation of the cyclohexene oligomers Bromination of the cyclohexene oligomer mixture. A solution of Bri in EtaO was added dropwise to a solution of the oligomers in EtiO. The solution was initially yellow upon addition of the Bri, and then it became colorless. Eventually the color no longer faded. The solvents were removed to obtain a clear oil. This oil was redissolved in EtiO and analyzed by the usual GC/MS methodology. The resulting GC trace and M S spectra resembled the data for the unmodified cyclohexene oligomers. There was no evidence for any bromination of the oligomers. Thin layer chromatography. Thin layer chromatography was attempted using Aluminium oxide 60 F254 neutral plates and Silica gel 60 F254 plates (both obtained from E M  Science). In both cases the plates were cut to 2.5 x 8.5 cm and the mobile phase was pentane. Development times ranged from 1 to 12 min. In all cases, the mixture of cyclohexene oligomers moved just behind the solvent front, without any separation into distinct bands. Silver nitrate impregnated silica T L C plates were prepared based on published methodology.*° The siUca plates (2.5 x 8.5 cm) were developed with a solution of AgNOs (4 g) in water (10 mL) and then dried with a heat gun. The plates darkened as they were dried. Subsequent T L C analyses of the cyclohexene oligomer mixture and of the distilled dimers, with hexanes as the mobile phase, produced no movement of the oligomers. Selective solvation. Fractionation of the cyclohexene oligomers was attempted using selective solvation in acetonitrile and in methanol. Cyclohexene oligomers (0.100 g) were placed in a preweighed vial. Acetonitrile (10 mL) was added, and the vial was shaken vigorously and left overnight. The solution was separated from the undissolved oil, the acetonitrile was removed and the residues were analyzed by standard GC/MS methods. A similar test was conducted with methanol, except that the solvent was added in two amounts ( 2 x 5 mL) and each part analyzed separately. Representative compositions of the fractions are summarized in Figure 2.11. Acetone was also tested without success. HPLC analyses. H P L C analyses of the cyclohexene oligomers were vindertaken with the help of Brian Mendelsohn from the group of Dr. Marco Ciufolini. The full cyclohexeneoligomer mixture was run on an analytical H P L C column with a water/acetonitrile solvent gradient. The resulting trace showed many peaks, but none could be assigned. Further tests were done on a sample containing 84.6% cyclohexene trimers (as obtained by distillation). Eleven samples taken over 50 minutes were coUected from a prep-scale H P L C machine. Analysis of the fractions by *H N M R spectroscopy and GC/MS standard methods gave no matches for the expected cyclohexene oligomers. A similar coUection of fractions from an analytical HPLC run, with analysis by G C / M S , also gave no positive resuhs.  2.4.7 Bulk separation: Vacuum distillation of the cyclohexene oligomer mixture The distillation of the cyclohexene oligomer mixture (5.3 g) was set up with a silicone oil bath and a Vigreux distillation column packed with glass beads. The system was placed under dynamic vacuum, and the mixture was heated with stirring. A first fraction (1.9 g) distilled at 55 °C, and a second fraction (1.0 g) distilled at 116-120 °C. The oil remaining in the still pot constituted a third fraction. The viscosity of the still pot contents was noticeably greater than that  of the distilled fractions, and that of the second fraction greater than the first fraction. A small amount of each fraction was dissolved in Et20 and analyzed by G C / M S utilizing a method analogous to that employed for the oligomer mixture. The percentage composition of each fraction was determined by integration of the G C trace, and the results are illustrated in Figure 2.3. The first fraction contained 100% dimers of cyclohexene. The second fraction contained 3.4% dimers, 94.3%) trimers, and 0.5% tetramers. The third fraction (still pot) contained 19.7% trimers and all the higher oligomers {56.2% tetramers and 20.9% pentamers as detected by G C ) . It is worth noting that the higher cyclohexene oligomers were still thermally stable at 245 ° C , the final temperature of the oil bath. It should also be noted that exposure of the cyclohexene oligomers to air for extended periods of time (weeks) caused changes to the oligomers. The samples tumed light yellow and the viscosities increased. G C / M S analysis of a trimer sample exposed to air gave mass spectra with parent masses increased by 18 or 20 amu {m/z = 264, 260, 258) in addition to the masses typical of the expected trimers {m/z = 246, 244).  2.4.8 Identifícation of the major cyclohexene dimer A sample of previously distilled cyclohexene dimers (0.435 g , -0.5 mL) was injected into a flask containing Wilkinson's catalyst (0.045 g) dissolved in benzene (15 mL) under a hydrogen atmosphere. The solution was stirred vigorously ovemight. The solvent was then removed in vacuo, and the residue filtered through a Florisil column (3 x 1 cm) with pentane. The organic products were collected and analyzed by N M R spectroscopy (Figures 2.6 and 2.7). The hydrogenation of the dimers was complete, with one major product present. This product was identified as cyclohexylcyclohexane. A sample of previously distilled cyclohexene dimers (0.91 g, -0.95 mL) was injected into a flask containing Wilkinson's catalyst (0.103 g) dissolved in benzene (20 mL) under a hydrogen atmosphere. The solution was stirred for 3 h. The solvent was removed in vacuo, and the residue filtered through a Florisil column ( 2 x 3 cm) with pentane. The organic products were collected and analyzed by N M R spectroscopy (Figures 2.8 and 2.9). The hydrogenation of the dimers was not yet complete, and two major products were present. The first product was identified as cyclohexylcyclohexane by comparison of its representative ' H and ' ^ C N M R peaks to the fiíUy hydrogenated sample {vide suprd). The second product was identified as 3-cyclohexylcyclohexene, based on the olefinic peaks. It was also  identifíed as the major mono-unsaturated dimer present in the original dimer mixture, based on comparisons between the N M R spectra of the original dimer mixture and those of the identifíed 3 -cyclohexylcyclohexene.  (  b  c d  n  Characterization data for the mixture of cyclohexylcyclohexane and 3cyclohexylcyclohexene. ' H N M R (600 M H z , CeDé, 25 °C) 5 0.9-1.06 (2H, b; 2H, /; IH, a), 1.08-1.24 (IH, d; IH, n; I H , h- 2H, c; 2H, m), 1.27-1.34 (IH, k), 1.44-1.51 (lH,y), 1.57-1.77 (IH, k'; IH, d'; 1 H , « ' ; l H , é ' ; l H , /'; IH, y'; 2H, c'; 2H, m'), 1.88-1.94 (IH,  2H,/), 5.63-5.67  (IH, e), 5.70-5.74 (IH,/). Almost all signáis were complex, overlapping multiplets whose J valúes could not be analyzed or determined. The signáis arising from e and/were an A B pair; however, it was clear that the splitting patterns were different from each other and consistent with the structure of 3-cyclohexylcyclohexene. ' ' C j ' H } N M R (150 M H z , CgDé, 25 °C) 5 22.7 (/•), 25.8 (O, 26.1 {k), 21 A, 27.2 (three peaks; m, m', n), 27.3 (c and d), 30.1 (/ or /'), 30.5 (b), 30.6 (/ or /'), 41.3 (g), 43.0 (h), 43.8 (d), 127.4 (/), 131.2 (e). Peak assignments were made using ' H - ' H COSY, ' H - ' ^ C H M Q C and ' H - ' ^ C H M B C data (600 MHz). The relative ratio of cyclohexylcyclohexane to 3-cyclohexylcyclohexene in the mixture was approximately 1:2 based on the ' H and '^c N M R data. N M R data (600 MHz) for the original dimer mixture revealed that 3cyclohexylcyclohexene was the major unsaturated dimer. The other unsaturated dimers present were not identifiable. A very small amount of cyclohexylcyclohexane was also present in the mixture.  2.5 References (1)  Introduction to Spectroscopy, 2"'' ed., D. L. Pavia, G. M . Lampman and G. S. ICriz, Saunders College Publishing, Orlando, 1996.  (2)  Christensen, N . J.; Hunter, A. D.; Legzdins, P. Organometallics 1989, 8, 930-940 and references cited therein.  (3)  a) Farona, M . F.; Tsonis, C. P. J. Chem. Soc, Chem. Commun. 1977, 363-364. b) Wazeer, M . I. M . ; Tsonis, C. P. Polym. Bull. 1984,12, 475-479.  (4)  Moulijn, J. A . ; van de Nouland, B. M . React. Kinet. Catal. Lett. 1975, 3, 405-408.  (5)  Giezynski, R.; Korda, A . J. Mol. Catal. 1980, 7, 349-354.  (6)  Debad, J. D.; Legzdins, P.; Rettig, S. J.; Veltheer, J. E. Organometallics 1993,12, 27142725.  (7)  Debad, J. D.; Legzdins, P.; Lumb, S. A.; Rettig, S. J.; Batchelor, R. J.; Einstein, F. W. B. Organometallics 1999,18, 3414-3428.  (8)  Debad, J. D.; Legzdins, P.; Lumb, S. J. Am. Chem. Soc 1995,117, 3288-3289.  (9)  a) Legzdins, P.; Rettig, S. J.; Sánchez, L. Organometallics 1988, 7, 2394-2403. b) Wada, K.; Pamplin, C. B.; Legzdins, P.; Patrick, B. O.; Tsyba, I.; Bau, R. J. Am. Chem. Soc 2003,125, 7035-7048.  (10)  L i , T.-S.; L i , J.-T.; L i , H.-Z. J. Chromatogr. A, 1995, 715, 372-375.  C h a p t e r 3. R e a c t i o n s o f C p * M o ( N O ) ( C H 2 C M e 3 ) 2 w i t h Cyclic  Olefins^  A versión of this chapter has been pubhshed. Graham, P. M . ; Buschhaus, M . S. A . ; Legzdins, P. J. Am. Chem. Soc. 2006, J28, 9038-9039, and Graham, P. M . ; Buschhaus, M . S. A.; PampHn, C. B.; Legzdins, P. Organometallics, 2008, 27, 2840-285 L Reproduced in part with permission from the above articles. Copyright 2006 and 2008 American Chemical Society.  3.1 Introduction Insight into the specific mechanism of the Cp'W(NO)(CH2CMe3)2 systems with regards to their reactivity with cyclic olefms can be gained by examining the analogous molybdenum system. The complex Cp*Mo(NO)(CH2CMe3)2 (3) forms a transient alkylidene intermedíate in a marmer analogous to the related tungsten system, except that the reaction proceeds in solution at room temperature.' Like the tungsten system, 3 activates a range of aryl and alkyl C-H bonds.' In the presence of cyclic olefins, the reactivity of 3 should parallel the tungsten system. However, since the reaction temperature is lower, stable organometallic complexes can be isolated and characterized. This chapter describes the investigations of the reactivity of Cp*Mo(NO)(CH2CMe3)2 (3) with the cyclic olefms and delineates the general mechanism that explains this reactivity. At this juncture, the valuable work of Dr. Peter Graham must be acknowledged. His work includes the investigations involving cycloheptene and cyclooctene, as well as many of the investigations involving cyclopentene. Data for the reactivity of cyclohexene, some of the reactivity of cyclopentene, and all of the solid-state molecular structures are the responsibility of the author. 3.2 Results and Discussion  3.2.1 General reaction mechanism for the molybdenum system Reaction of 3 with cyclic olefins leads to a series of complexes along the reaction pathway depicted in Scheme 3.1. Varying the cyclic-olefm ring size allows isolation of the key species in the mechanism. A five-membered ring (cyclopentene) does not proceed as far along the reaction pathway as a six-membered ring (cyclohexene), which in tum proceeds further than an eight-membered ring (cyclooctene).  ,MoO  trans-metallacycle  ,Mo\ H  O  coupled organic  TI -allyl hydride  O  TI -diene  Tl^-allyl hydride  The first step on the reaction pathway is the well-established formation of the alkyUdene intermediate' in solution under ambient conditions and its subsequent reaction with a reagent molecule to form a molybdenacyclobutane. This initial 2 + 2 addition results in a metallacycle, with the two formerly-olefmic hydrogens on the same side of the ring. Rearrangement of this cismetallacycle, via an rj^-allyl-hydride complex, forms the trans-metallacycle, where the two hydrogens are now on opposite sides of the ring. The trans-metallacycle can revert to the r]^allyl hydride, which is proposed to form a transient ri'-allyl hydride. This species then loses dihydrogen (H2), resulting in the formation of an r|''-diene complex. Finally, under appropriate reaction conditions, the coupled organic ligand can be released from the metal center. Each step of the general mechanism is supported by data from at least one of the four ring-sizes examined. The details of the interesting results of each cyclic olefin's reactivity are examined in tum, together with the reasoning supporting the formulation of each major step of the general mechanism.  3.2.2 Reactivity of 3 with cyclopentene: The cis-metallacycle 7 The first step of the general mechanism involves the formation of a molybdenacyclobutane complex by a 2 + 2 cycloaddition of the alkylidene intermediate with the cyclic olefin. Evidence for this step is the formation of 7 from the reaction of 3 with cyclopentene, as shown in Scheme 3.2. At 20 °C for 26 h, orange-colored 7 forms exclusively, and the isolated product may be recrystallized from pentane/EtaO.  Scheme 3.2  3  7  Complex 7 has been ñiUy characterized by conventional methods, including X-ray crystallography. The single-crystal X-ray diffraction analysis (Figure 3.1) confirms the metallacyclobutane structure. The plañe defmed by C l , C2, C6 and M o is almost fíat, as evidenced by the C 2 - C l - C 6 - M o l torsión angle of 0.34(14) degrees. The protons H l and H2 are positioned cis, situated on the same side as each other relative to the ring. Their exact positions are crystallographically fixed. The final R valué for this structure is 2.0%, attesting to the quality of the data.  Figure 3.1 Solid-state molecular structure of 7 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): Mo(l)-C(2) = 2.0599(19), Mo(l)-C(l) = 2.3394(19), Mo(l)-C(6) = 2.1050(18), C(l)-C(2) = 1.693(3), C(l)-C(6) = 1.561(2), Mo(l)-C(2)C(l) = 76.43(9), Mo(l)-C(6)-C(l) = 77.74(10), C(6)-C(l)-C(2) = 120.42(15), C(2)-Mo(l)-C(6) = 85.41(7), Mo(l)-C(2)-C(3) = 129.43(15), C(6)-C(l)-C(5) = 113.72(16), C(2)-C(l)-C(6)-Mo(l) = 0.34(14).  The diagnostic ' H and '^C N M R signáis of 7 are of interest. Specifically, in the ' H N M R spectrum in CeDg, the a-proton of the metallacycle (H2) gives rise to a downfíeld resonance at 7.94 ppm, while the adjoining p-proton (Hl) gives rise to an upfield resonance at -0.05 ppm. The '^C N M R data confírm the metallacyclobutane structure, with signáis for the two carbons bonded to the molybdenum (C2 and C6) occurring at 154.2 ppm and 122.7 ppm, respectively. The third carbón in the metallacycle, C l , gives a signal at 20.4 ppm. Similarities can be traced between the related molybdenum complexes derived fi-om the various cyclic-olefm ring sizes. The diagnostic ' H and '^C N M R signáis for the cis- and transmetallacycles and the ri'*-diene complexes of molybdenum are summarized on the following page in Table 3.1.  3.2.3 Further reactivity of 7: The allyl-hydride complex 8 At room temperature in the solid-state, complex 7 shows further reactivity to form the allyl-hydride complex 8, as shown in Scheme 3.3. This reactivity also proceeds in polar solvents such as THF. However, in non-polar solvents such as C^De the reaction is very slow, and heating leads only to eventual decomposition.  Scheme 3.3  7  8  H N M R spectroscopic analysis of 8 shows signáis for the allylic protons at 1.61 ppm and 1.59 ppm. The signal ascribed to the metal hydride appears at -2.04 ppm. The allylic carbons give signáis in the '^C N M R spectrum at 123.2 ppm for the central quatemary carbón, and at 81.1 ppm and 80.0 ppm for the adjacent outer carbons.  Table 3.1 Diagnostic ' H - and ' ^ C - N M R data for compounds 7-10 and 13-17.  O'. o  i  Mo  neat or cyclohexane  metallacycles  'H (ppm)  " C (ppm)  Hl- -0.05 (ddd)  Cl;  H2: 7.49 (dd)  C2: 122.7  H6: 3.27 (d)  C6: 154.2  Hl -0.46 (dd)  cis-metallacycles  diene  allyl hydride  "C(ppm)  H2: 1.61 (brs)  C2  81.1  H6: 1.59 (brs)  C6  80.0  Cl 7  Mo 1 5  123.2  hydride; -2.04 (s)  Cl C2 110.2  H8  C8 155.5  O  'H (ppm)  "C(ppm)  trans-metallacycles  H l -0.65 (m)  Cl  H2 O.I (ddd)  C2  H7, 3.91 (d)  C7  13  dienes  'H(ppm)  " C (ppm)  C l ; 115.8 17  Mo—I •  16  7r 1 6  Hl;-0.28(dddd)  Cl  H2; 4.43 (ddd)  C2 130.9  H8: 2.70 (d)  C8 138.2  15.2  H l -0.18 (ddd)  Cl  H2 4.10 (ddd)  C2 121.5  H9  C9 152.3  4.31 (d)  'H(ppm)  20.4  H2 5.66 (dd) 4.52 (d)  ,  n = 2-4  Mo  14 O  ,Mo  15  o  14.9 ,Mo-^-"'~  ^  C2; 85.9  H3; 3.73 (ddd)  C3; 76.9  H7; 1.70 (s)  C7; 94.1  Cl;  2....^4  ,N"  H2; 1.12 (d)  3 4  8 7  10  H2 0.21 (d)  -  C2; 93.6  H3  3.21 (ddd)  C3;^ 82.2  H8  1.40 (s)  C8; .81.1  Cl;  21.3  H2 0.72 (d)  C2;  85.2  H3  3.06 (ddd)  C3: 78.0  H9  1.67 (s)  C9; 96.7  1 í  The solid-state molecular structure of 8, shown in Figure 3.2, confirms the allylic nature of the organic ligand, as well as the position of the hydride. The Mo-C bond lengths in the allyl moiety range from 2.33 to 2.36 Á. The position of the hydride (HOl) is located and refined on the basis of residual electrón density. This is possible because of the excellent quality of the X ray data. The final R valué for the structure is 2.4%.  Figure 3.2 Solid-state molecular structure of 8 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): Mo(l)-C(2) = 2.3581(18), Mo(l)-C(l) = 2.3433(17), Mo(l)-C(6) = 2.3272(17), C(l)-C(2) = 1.397(3), C(l)-C(6) = 1.422(2), C(6)-C(l)C(2) - 121.17(17), Mo(l)-C(2)-C(l)-C(6) = 55.56(15).  Compared to the parent cis-metallacycle (7), the cyclopentyl ring in 8 has shifted position, and as a result the protón on C2 is now pointed upward towards the Cp* ligand. This configuration around C2 looks similar to that of the trans-metallacycle (vide infra), and therefore  an r\ -allyl-hydride type complex is proposed as an intermedíate species between the cis- and trans-metallacycles. The transformation from the t]'-allyl hydride to the trans-metallacycle is believed to be reversible, such that the ri^-allyl hydride is also the intermedíate en route to the formation of the t]'-allyl hydride and the diene complexes. Cyclopentene is the only cyclic olefin to afford direct evidence for an rj •'-allyl-hydride species. It seems that the trans-metallacycle is inaccessible from 8, perhaps due to the small size of the ring and the strain that such a transformation would impose. A l l of the larger cyclic olefíns can form the trans-metallacycle. They do not produce an observable allyl hydride although such a species is the presumed intermedíate between the cis- and trans-metallacycles and between the metallacycles and the diene complex. These transformations occur quickly on the N M R time scale, and ' H N M R monitoring reveáis no trace of the r|'-allyl-hydride intermediate.  3.2.4 Reactivity of 3 with cyclooctene: The trans-metallacycle 9 The formation of a trans-metallacycle complex is best exemplified by the reaction of 3 with cyclooctene to form 9 (Scheme 3.4). The reaction occurs at room temperature, in a mixture of cyclooctene and cyclohexane, and yields upon work-up orange crystals of 9 as the major organometallic product.  Scheme 3.4  cyclohexane 3  9  The solid-state molecular structure of 9, pictured in Figure 3.3, shows the molybdenacyclobutane motif, with the C l , C2, C9 and M o plañe again being almost fíat. The C 9 - C l - C 2 - M o l torsión angle is 3.42(17) degrees. The structure also clearly shows the relative positions of H l and H2 to be trans to each other, on opposite sides of the ring. The configuration  around C2 has the protón H2 pointed upward towards the Cp* ring. This differs from the solidstate structure of 7 (a cis-metallacycle) but is similar to the solid-state structure of 8 (an allyl hydride). In this structure, H l , H2 and H9 have been refined to the best positions, and the final R valué is 2.5%.  Figure 3.3 Solid-state molecular structure of 9 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): Mo(l)-C(2) = 2.071(2), Mo(l)-C(l) = 2.340(2), Mo(l)-C(9) = 2.070(2), C(l)-C(2) = 1.608(3), C(l)-C(9) = 1.635(3), Mo(l)-C(2)-C(l) = 77.90(11), Mo(l)-C(9)-C(l) = 77.37(11), C(9)-C(l)-C(2) = 119.47(17), C(2)-Mo(l)-C(9) = 85.13(8), Mo(l)-C(2)-C(3) = 126.31(16), C(9)-C(l)-C(8) = 107.25(17), C(9)-C(l)-C(2)-Mo(l) = 3.42(17).  Characterization of 9 by H N M R spectroscopy reveáis diagnostic signáis for the aproton (H2) of the metallacycle at 4.10 ppm and for the P-proton ( H l ) at -0.18 ppm. In particular, compared to the cyclopentene-derived cis-metallacycle (7), the a-proton signal of the trans-metallacycle has shifted upfield. '^C N M R data again confirm a metallacyclobutane structure. The signáis for the two carbons bonded to the molybdenum (C2 and C9) occur at 121.5 ppm and 152.3 ppm, respecüvely. The third carbón in the metallacycle, C l , gives a signal at 14.9 ppm.  3.2.5 Further reactivity of 9: Formation of ''-diene 10 upon heating Heating 9 in cyclohexane at 50 °C for 48 h results in the formation of the yellow r(*~ diene complex 10 (Scheme 3.5). Complex 10 is analogous to the previously discovered cyclohexene-derived ri''-diene complex, which is discussed later in Section 3.2.9.  Scheme 3.5  9  10  Evidence for the concomitant reléase of H2 during the conversión of 9 to 10 is provided by an NMR-scale experiment. Upon thermolysis of 9 at 50 °C in cyclohexane-£/y2 and cyclohexene, a small singlet appears at 4.54 ppm in the ' H N M R spectrum, in addition to the signáis expected for complex 10. This singlet indicates the formation of H2. The small size of the signal is a result of the low solubility of H2 in cyclohexane-íi;!The solid-state molecular structure of 10 is disordered in the cyclooctenyl ring, specifically carbons C5, C6 and C7. This disorder can be modeled with two orientations of 40% and 60% occupancy, and Figure 3.4 shows the major orientation. Constraints have been used to  keep the related bond lengths in the disordered región equivalent. The three disordered carbons have been refined isotropically, while all other non-hydrogen atoms have been refined anisotropically. The protons near the metal center, H2, H3 and H9, have been refined isotropically. Unfortunately, due to the disorder in the structure, the final R valué is high at 6.3%.  Figure 3.4 Solid-state molecular structure of 10 with 50 %> probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): Mo(l)-C(l) = 2.303(6), Mo(l)-C(2) = 2.182(6), Mo(l)-C(3) - 2.344(7), Mo(l)-C(9) = 2.362(6), C(l)-C(2) = 1.429(9), C(2)-C(3) 1.426(9), C(l)-C(9) = 1.414(10), C(l)-C(2)-C(3) = 123.4(6), C(2)-C(l)-C(9) = 114.2(5), C(3)C(2)-C(l)-C(9) = 124.6(7), C(l)-C(2)-C(3)-C(4) = 43.3(10), C(2)-C(l)-C(9)-C(10) - 177.5(6).  The ' H N M R data for 10 show diagnostic signáis for H3 at 3.06 ppm, for H2 at 0.72 ppm, and for H9 at 1.66 ppm. The '^C N M R data indícate that C l , the quatemary carbón attached to the metal center, gives rise to a signal at 121.3 ppm. The other three tertiary carbons that interact with the metal center give signáis at 96.7 ppm (C9), 85.2 ppm (C2), and 78.0 ppm (C3). The interaction between the four carbón atoms of the organic ligand and the metal center in complex 10 is best described as a conjugated diene distorted away from ideal parameters by the metal. In the " C N M R data the similarity in shift for the signáis due to the tertiary carbons interacting with the metal center, between 80 and 100 ppm, suggest a diene ligand in which the interactions are all equivalent. In contrast, alkyl-type linkages in the metallacyclic complexes give rise to signáis from 110 to 160 ppm (Table 3.1). Solid-state metrical parameters for 10 and for its cyclohexene relative {vide infra) suggest a diene by virtue of the C-C bond lengths, which range in length from 1.40 A to 1.45 A. The bond lengths are equivalent (within error) between all of the carbons that interact with the metal center, and are similar to other known Cp*Mo diene complexes.^^ The torsión angles in 10 indícate that the four carbons do not lie in a perfect plañe. The torsión angle C2-C1-C9-C10 is 177.5°, which is cióse to 180° and therefore indicates the near planarity of those carbón atoms. The torsión in the C4-C3-C2-C1 bonds is 43.3°, somewhat twisted from the expected 0° planarity. Finally the torsión of C3-C2-C1-C9 is 124.6°, significantly twisted from the 180° planarity that would be expected in a conjugated diene without any other interactions to affect it. The cyclohexene-derived complex gives solid-state parameter valúes similar to those of 10. These torsión angles, particularly the central one of 124.6°, can be rationalized as the diene twisting in order to bond to the metal center. A previously reported CpMo transdiene complex gives a similar central torsión angle of 124.8(4)°.'" Finally, the V N O stretching frequencies for the t)''-diene complexes range from 1589 cm"' to 1598 cm"'. These valúes are similar to those of the cis- and trans-metallacycles ( V N O = 1586 cm"'to 1588 cm"), and also similar to other known Cp*Mo diene complexes.  3.2.6 Reactivity of 10: Trapping the ri^-diene complex 11 with pyridine In the presence of neat pyridine at 100 °C for 19 h, complex 10 forms a pyridine-trapped r|^-diene complex (11), as depicted in Scheme 3.6.  Scheme 3.6  ^  O  y \  10  \ /  pyridine 100°C, 19h  O  11  The solid-state molecular structure of 11 (Figure 3.5, next page) confírms that the double bond of the cyclooctenyl ring remains coordinated to the metal center. The C1-C2 bond length is 1.44 A. The C8-C9 double bond is 1.25 A, while the C1-C8 single bond is 1.48 A. The elongation of the C1-C2 bond, such that the length compares more to the single bond than to the noncoordinated double bond, is attributable to back-bonding interaction with the metal. The protons of the olefin bonds ( H l , H2 and H9) have been refined isotropically based on the residual electrón density. The final R valué for the structure is 3.4%. The diagnostic ' H N M R signáis for 11 include the singlet for the protón (H9) of the noncoordinated olefin at 4.80 ppm, and the multiplets for H l and H2 of the coordinated olefin at 1.90 ppm and 1.87 ppm, respectively. Similarly, the '•'C N M R data distinguish the coordinated and noncoordinated olefíns. C l and C2 give rise to signáis at 77.1 ppm and 61.9 ppm, respectively, indicating their interaction with the metal center. C9 gives a signal consistent with a free olefín at 129.3 ppm, and the signal for quatemary C8 can be observed with an H M B C N M R experiment at 145.1 ppm.  Figure 3.5 Solid-state molecular structure of 11 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): Mo(l)-C(l) = 2.273(3), Mo(l)-C(2) = 2.212(3), Mo(l)-N(2) = 2.192(2), C(l)-C(2) = 1.437(4), C(l)-C(8) = 1.484(3), C(8)-C(9) = 1.345(4), C(2)-C(l)-C(8) = 130.1(2), C(l)-C(8)-C(9) = 117.5(2), C(2)-C(l)-C(8)-C(9) 161.6(3).  3.2.7 Reactivity of 10: Reléase of the coupled organic ligand 12 The coupled ligand can be released by heating 10 in cyclopentene at 70 °C for 3 days (Scheme 3.7). The organic product, 3-neopentylidene-cycIooctene (12), is isolated as a colourless oil using column chromatography. As expected, 12 gives rise to three olefmic signáis in its ' H N M R spectrum in CeDe, at 6.21 ppm, 5.57 ppm and 5.37 ppm. The '^C N M R data show four olefmic carbón signáis at 144.5 ppm, 140.3 ppm, 138.6 ppm and 124.8 ppm.  No discrete organometallic molybdenum products have been identified in the reaction mixture generated by the thermolysis of 10, even though a variety of different solvents were tested in hopes that one might stabilize or trap such products. Intuitively, the loss of the coupled ligand suggests that the Cp*Mo(NO) fragment would remain. Evidently that fragment undergoes decomposition by unknown pathways. Interestingly, when cyclohexene is used as the solvent for the thermolysis reaction, small amounts of the related cyclohexene oligomers are detected in the product mixture. Allylbenzene, when used as the thermolysis solvent, yields an allylbenzene dimer product. These organic products hint at a link with the previously observed tungsten oligomerization reactivity. The implications of this are further developed in Chapter 5 when the tungsten mechanism is considered.  3.2.8 Sequential observation of metallacycles and an T| -diene complex: Reactivity with cycloheptene Cycloheptene reacts with 3 to form the cis-metallacycle 13, the trans-metallacycle 14, and the r|''-diene complex 15 sequentially (Scheme 3.8). The reactions occur in a mixture of cycloheptene and cyclohexane, and notably the presence of the cyclic alkane does not interfere. After 6 hours reaction time, complex 13 is characterized by ' H - and ' ^ C - N M R spectroscopy as a mixture with 14. After 20 hours reaction time, complex 14 can be isolated as dark red crystals and ftilly characterized. Finally, when the reaction is complete (after 312 h), complex 15 is  isolated as a yellow solid. The formation of 15 is expedited by heating (50 °C), and the reaction time is thus reduced from many days to 20 hours.  Scheme 3.8  3  O  O  rt  13  14  15  Monitoring the reaction by H N M R spectroscopy at room temperature over 13 days demonstrates the sequential formation of these products, as illustrated in Figure 3.6. The peaks due to the starting material 3 disappear as peaks arising from compound 13 appear (t = 6 h). Then peaks arising from compound 14 begin to appear (t = 6 to 20 h), and then finally those of 15 (t = 20 h). By the end of the reaction time, only compound 15 is present in the solution (t = 312 h). Because each of the complexes has been individually characterized, the peak assignments in the reaction mixture are defínitive. This sequential observation of the cismetallacycle, the trans-metallacycle and the ri'^-diene complexes elegantly supports the formulation of the general mechanism (Scheme 3.1).  0,1 h 6h 20 h 30 h 66 h 312 h  0 4.5  4.0  3.5  D  ppm  Figure 3.6 Time lapse ' H N M R data for the reaction of 3 in cycloheptene. Selected diagnostic signáis in the 2.7 to 5.0 ppm región, t = 0.1 to 3 1 2 h (cycloheptene/cyclohexane-¿//7, rt). Peak labels: 1 = 3 (methylene H ) , A = 13 ( H 8 ) , C = 14 ( H 2 and H 8 ) , D = 15 ( H 3 and H 7 ) .  3,2.9 Reactivity of 3 with cyclohexene: The t) -diene complex 16 Cyclohexene in reaction with 3 produces the r]''-diene complex 16 (Scheme 3.9). The reaction proceeds at room temperature to produce yellow 16, which is analogous to the r|''-diene complex 10 isolated from the reaction with cyclooctene. On the basis of spectroscopic evidence, the metallacycle 17 is proposed as an intermediate in the reaction pathway (vide infra).  Scheme 3.9  3  17  16  Figure 3.7 displays one view of the solid-state molecular structure of 16. The X-ray crystallographic data for 16 indícate that the coupled organic ligand and the nitrosyl ligand are disordered in two mirror-image orientations. As a result, atoms which overlap in the two orientations (i.e. the N atom and C2) and atoms lying roughly on the plañe between the two orientations (i.e. C8 and C U ) have been refined isotropically instead of anisotropically. Further, restraints have been applied to maintain similar geometries for the two forms of each related segment, namely the rerí-butyl groups, the cyclohexyl rings, and the nitrosyl ligands. The protons on C2 and C3 have been refined isotropically, while all other protons have been included in fixed positions. Despite the care necessary to solve the structure, the final R valué is a reasonable 2.9%.  Figure 3.7 Solid-state molecular structure of 16 with 50 %> probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): Mo(l)-C(7) = 2.299(4), Mo(l)-C(l) = 2.233(4), Mo(l)-C(2) - 2.115(9), Mo(l)-C(3) = 2.301(5), C(3)-C(2) = 1.403(8), C(2)-C(l) = 1.449(8), C(l)-C(7) = 1.427(6), C(3)-C(2)-C(l) = 117.0(6), C(2)-C(l)-C(7) =114.9(5), C(3)C(2)-C(l)-C(7)= 126.5(6).  The H N M R data of interest for 16 include the signáis due to the protons nearest the metal center. Specifically, the signal for H 3 appears at 3.74 ppm, that of H 7 at 1.70 ppm, and that of H 2 at 1.12 ppm. In the '^C N M R spectrum, the signáis of C l , C 2 , C 3 and C 7 appear at 115.8 ppm, 85.9 ppm, 7 6 . 9 ppm and 94.1 ppm, respectively. This corresponds to a quatemary carbón and three tertiary carbons interacting with the metal center. When the yellow compound 16 is isolated by column chromatography on neutral alumina, a second minor product can be seen as a blue-green band on the column. This band elutes after and slightly overlaps the yellow band. The blue product thus obtained is more soluble in pentane than is 16, and 16 can be selectively precipitated from a mixture of the two products. Regrettably, the minor blue product has not been identifíed. When it is redissolved in cyclohexene at room temperature in the glovebox, it converts to 16 in days, as confírmed by ' H N M R spectroscopy. As a result of this conversión, it is extremely difficult to obtain N M R data for the blue product because 16 always contaminates the sample and dominates the N M R signáis, such that the peaks arising from the blue product can not be identified. Complex 16 is definitely the main and the most thermodynamically stable organometallic product. In comparison with the general reactivity of the molybdenum system, it seems unlikely that the blue product represents a metallacyclobutane compound, since these types of compounds are orange for other ring-sizes.  3.2.10 Evidence for the metallacyclobutane intermediate 17 in the cyclohexene reaction Evidence for the existence of at least one metallacyclobutane intermediate comes from experiments using UV/vis spectroscopy and ' H N M R spectroscopy to monitor the progress of the reaction of 3 with cyclohexene to ultimately form 16. Unfortunately, typically inert solvents such as cyclic alkanes interfere with the cyclohexene reaction, unlike the reactivity of 3 with cycloheptene (vide supra). Therefore, evidence for 17 is obtained in more roundabout ways. The UV/vis spectroscopy experiment monitoring the reaction of 3 with cyclohexene shows the formation of intermediate product(s) before the formation of the final product 16. At concentrations of ca. 1 x lO'^ M compound 3 in cyclohexene gives a spectrum with three main peaks at 2 2 8 rmi, 2 9 2 imi and 4 8 2 nm. A n authentic sample of 16 under similar conditions gives a spectmm with a main peak at 2 3 0 nm, and a peak at 2 8 6 rmi with a strong shoulder at 3 3 0 nm. Monitoring the reaction over time, initially with 10 minute intervals and then with 15 minute intervals for the first day and much longer intervals for the following three days, showed  the consumption of 3 and the formation of 16. The data also show the presence of one or two intermediate species whose diagnostic peaks fall between those of 3 and 16. These diagnostic peaks seem to correlate with the observed color changes in the solution, from red to orange to yellow. As the 482 nm peak arising from 3 decreases, a new peak grows in at approximately 414 nm, with an isosbestic point forming at 446 nm. A second isosbestic point later forms at 452 nm (although not as clearly as the first point). Figure 3.8 shows the UV/vis data collected over the fírst day. The major peaks decrease at 290 nm and at 482 nm, and increase at 414 nm. Figure 3.9 contains an expansión of the first day's data to show the isosbestic points. The growing peak at 414 nm represents the intermediate species. In the later scans (2"^* to 4* day) the peak at 330 nm arising from 16 grows in to domínate the spectra.  0.800  0.700 ^  0.000 260  310  360  410  460  510  560  Wavelength (nm)  Figure 3.8 UV/vis data monitoring the reaction of 3 with cyclohexene, first day only (wavelengths 260-580 nm, 26 scans, t = 3 min to 572 min). Peaks decrease at 482 nm and 290 nm, and increase at 414 nm.  0.018  -I  ,  ,  ^  ,  ,  390  410  430  450  470  490  510  Wavelength (nm)  Figure 3.9 Expaiided UV/vis data from Figure 3.8, first day only (wavelengths 390-510 nm, 26 scans, t = 3 min to 572 min). Peak decreases at 482 nm and increases at 414 nm. The first isosbestic point is at 446 nm, and the second isosbesüc peak is at 452 nm.  Overall the UV/vis experiment demonstrates the existence of intermediate species in the reaction of 3 and cyclohexene. But it tells little about the nature of the species. For further insight monitoring by ' H N M R spectroscopy was used. Initial ' H N M R experiments designed to detect the metallacyclobutane intermediate did not work well. When direct monitoring of a solution of 3 in cyclohexene for 28 hours with a CóDó capillary as a reference was undertaken, the signáis of the solvent obscured the signáis of the organometallic species. A similar experiment using 10 equivalents of cyclohexene in CD2CI2 did not give the same reaction products, and was therefore useless in regards to elucidating reaction intermediates. The reaction of 3 in cyclohexene can be tracked successñiUy by setting up the reaction and withdrawing 1 mL aliquots over time. The volátiles are immediately removed from each  aliquot in vacuo to isolate the organometallic species, and the residue of each aliquot is analyzed by ' H N M R spectroscopy to identify the organometallic species present. To summarize, analysis of the ' H N M R data of the aliquots taken over time reveáis the loss of the starting material (3), and the formation of a new product (17). This in tum more slowly converts to 16. Eventually 17 is completely consumed and only 16 remains. Some traces of decomposition products begin to appear near the end of the reaction time. Also, traces of organic products appear throughout the reaction sequence (yide infra). Selected ' H N M R signáis attributed to 17 suggest a metallacyclobutane-type complex. These diagnostic signáis include singlets at 1.7 ppm (Cp*) and 1.2 ppm (CMes), and multiplets that intégrate for one protón each at 3.9 ppm (d), 0.1 ppm (ddd) and -0.7 ppm (m). The smaller multiplet signáis are reminiscent of the characteristic metallacycle peaks of 9 in their splitting pattems and in their chemical shifts (particularly those at 3.9 and -0.7 ppm). Table 3.1 compares the ' H N M R valúes for the various metallacycles, including 17. On the basis of comparisons to 9 and 14, complex 17 is identified as a trans-metallacycle.  3.2.11 Cyclohexene-derived organic products The reaction of 3 with cyclohexene also produces organic products in addition to 16. The presence of these organic products is generally detected in the cmde reaction mixtures by ' H N M R spectroscopy as small signáis around 5.7 ppm. This corresponds to the olefmic protons of cyclohexene oligomers, as observed initially in the tungsten system. The oligomeric organic products are separated from the organometallic products by column chromatography on alumina, and are recovered from the initial pentane or hexanes fraction. G C / M S analyses of the isolated oligomeric organic products show marked similarities to the cyclohexene oligomers formed in the analogous tungsten systems. Specifically, in the GC trace peaks appear at 1.4 min, 1.9 min, and between 9 to 11 min. The M S data reveal that these peaks correspond to cyclohexene dimers with unsaturation, a neopentyl unit coupled to cyclohexene, and a mixture of trimers and neopentyl coupled to dimers, respectively. The relative yields of cyclohexene oligomers are much lower for the molybdenum system compared to the tungsten system. ' H N M R spectra of the cmde reaction products for the Mo system have a large peak due to the Cp* ligand of 16 that is about 30 times larger than the peak at 5.7 ppm arising from the olefmic protons of the oligomers. In contrast, in the comparable spectra for the tungsten system the peaks due to the oligomer products obscure those of all other  species. The tiny isolated amounts of oUgomer are difficult to quantify in the M o system, while the tungsten system yields gram amounts. Generally, in the M o system there is a higher representation of oligomers with a coupled neopentyl unit in the GC trace. Average unsaturation levéis per oligomer are also higher in the molybdenum system compared to the tungsten system.  3.2.12 Reactivity of 16 vt'ith cyclohexene and with mixed solvents under thermolysis conditions Investigations into the further reactivity of 16 demónstrate that the complex is a precatalyst for the oligomerization of cyclohexene under appropriate thermolysis conditions. (Although, it should be noted, it is not a highly productive precatalyst compared to the tungsten system.) A prepared solution of 16 in cyclohexene left at room temperature for eleven days produces traces of oligomeric products, as detected by ' H N M R spectroscopy. In a similar solution heated to 50 °C for nine days oligomers form and 16 remains as the only detectable organometallic complex present. Heating at 70 °C for three weeks yields greater amounts of oligomeric products, but some unconsumed 16 still remains. Finally, heating at 100 °C for two days produces relatively large amounts of the cyclohexene oligomers with total consumption of 16. The organometallic decomposition products can not be isolated or identifíed. A comparison of the relative integration ratios of the Cp*-protons' signal and the olefínic protons' signal in the ' H N M R spectra shows a shift from 32:1 at rt to 2:1 at 50 °C to 1:10 at 70 °C. No ratio can be determined for the 100 °C sample. These ratios reflect the increasing production of oligomers and the consumption of 16 with increasing temperature. Further to the thermolysis experiments outlined above, two experiments run in parallel have explored the effect of an 'inert' solvent on the reactivity of 16 with cyclohexene. In the fírst reaction 16 was dissolved in neat cyclohexene, while in the second reaction it was dissolved in a 1:9 mixture of cyclohexene and cyclohexane. After both were heated at 100 °C for one day, the Solutions both darkened from yellow to yellow-brown. As expected, the analysis of the products from the fírst reaction showed that 16 remained and that oligomeric organic products had formed, comparable to the 70 °C reaction above. In contrast, the second reaction showed that 16 remained without any organic-product formation. The suspected decomposition products in both reactions suggested by the brown color could not be detected by *H N M R spectroscopy.  In summary, these reactions establish that 16 reacts under thermolysis conditions and can catalyze the oligomerization of cyclohexene. The presence of an 'inert' cyclohexane solvent inhibits formation of the organic oligomer products.  3.2.13 Reactivity of the CpMo system with cyclohexene CpMo(NO)(CH2CMe3)2 (4) reacts analogously to Cp*Mo(NO)(CH2CMe3)2 except that it is more temperature sensitive and forms the reactive alkylidene intermediate at sub-zero temperatures.'* A reaction between 4 and cyclohexene yields an r)''-diene product (18) directly analogous to 16. The structure of 18 was confirmed by ' H - and ' ^ C - N M R analyses, and by lowresolution M S . No oligomeric products were detected in the crude reaction mixture. Because oligomers of the cyclic olefíns are of primary interest, and because the reactivity does not seem to differ from the Cp*Mo system, further investigations were not pursued with this system.  3.2.14 Qualitative comparisons of the cyclic olefín substrates Cyclopentene forms a cis-metallacycle and eventually an r|^-allyl hydride. The small size of the ring prevents formation of the trans-metallacycle. Larger ring sizes (6-8) proceed readily to a trans-metallacycle. Cyclohexene reacts to form the r|''-diene complex the fastest at room temperature. Cycloheptene also forms the ri'^-diene species but the reaction takes many days to go to completion. Finally, cyclooctene needs additional heating to form the r)''-diene compound from the trans-metallacycle. To generalize, smaller ring sizes proceed along the reaction pathway outlined in Scheme 3.1 at a faster rate. However, the five-membered ring can not proceed beyond the allyl-hydride complex due to its small ring size.  3.2.15 Reactions of 3 with acyclic olefíns Acyclic a-olefms such as 1-hexene, 1,7-octadiene and allylbenzene react with 3 to yield ri''-trans-diene complexes, as illustrated in Schemes 3.10 and 3.11. These products correlate with the cyclic-olefin ri''-diene complexes, and the mechanism for their formation involves a similar double C-H bond activation with coupling of the organic ligands.  r  -R M CMe4  o  -H2  R-J  í  19 R = ''Pr 20 R = ( C H 2 ) 3 C H - C H 2  Scheme 3.11  -Ph ,Mo. ON ^ 7  CMe4  -H2  oN" Ph21  Interestingly, the allylbenzene reaction mixture also yields organic products, which have been identified as allylbenzene coupled to neopentylidene (22) and allylbenzene dimer (23). The allylbenzene dimer can be isolated by column chromatography on alumina. The olefínic protón signáis in the N M R spectrvim, a doublet at 5.72 ppm and a doublet of triplets at 5.50 ppm correlated by a large coupling constant of 15.9 Hz, confírm a confíguration of an intemal double bond with E stereochemistry, and this is the only configuration observed. The yield of 23 can be optimized at 83 equivalents of 23 formed when one equivalent of 3 is thermolysed in 550 equivalents of allylbenzene substrate at 100 °C for 18 h. This dimerization, together with the smaller amounts of cyclic-olefin oligomerization observed, provide a link between the molybdenum system and the catalytically active tungsten system. Any mechanistic explanation must be able to account for both the cyclic-olefin oligomerization and the allylbenzene dimerization.  3.3 Summary In summary, a general mechanism for the reaction of Cp*Mo(NO)(CH2CMe3)2 (3) with cyclic olefins has been elucidated. The reaction pathway can be summarized as a series of complexes, namely a cis-metallacycle, an t)•'-allyl hydride, a trans-metallacyle, and an r|''-diene species. Specific examples of these complexes are isolated as the products of the reaction of 3 with cyclic olefins of various ring sizes. In addition, some acyclic a-olefms will couple with the alkylidene to form r|''-diene complexes, and allylbenzene is catalytically dimerized. This dimerization and the formation of small amounts of oligomeric organic products from the cyclic olefins suggest a tie to the related tungsten system.  3.4 Experimental Procedures  3.4.1 General Methods A l l reactions and subsequent manipulations involving organometallic reagents were performed under anaerobic and anhydrous conditions either at a vacuum-nitrogen dual manifold or in an inert-atmosphere dry box. Pentane, benzene-í/g, diethyl ether, and tetrahydrofuran (THF) were all dried over sodium benzophenone ketyl and were freshly distilled prior to use. Cyclopentene, cyclohexene, cycloheptene, and cyclooctene were all purchased from Aldrich, and were dried over sodium benzophenone ketyl, distilled, and stored in resealable glass vessels. Cp*Mo(NO)(CH2CMe3)2 (3) was prepared according to the published procedure.'^ A l l other Chemicals were purchased from Aldrich and used as received. A l l IR samples were prepared as Nujol muUs sandwiched between NaCl plates, and their spectra were recorded on a Thermo Nicolet Model 4700 FT-IR spectrometer. N M R spectra were recorded at room temperature on Bruker AV-300, Bruker AV-400 or Bruker AV-500 Instruments using standard U B C pulse sequences with delay times of 1 sec for ID experiments and 1.5 sec for 2D experiments. A l l chemical shifts and coupling constants are reported in ppm and in Hertz, respectively. ' H N M R spectra were referenced to the residual protio isotopomer present in CeDe (7.15 ppm). '^C N M R spectra were referenced to CóDe (128 ppm). Where necessary, ' H - ' H C O S Y , ' H - ' ^ C H M B C , ' H - " C H M Q C , and '^C A P T experiments were carried out to correlate and assign ' H and '^C N M R signáis. Low-resolution mass spectra (El, 70 eV) were recorded by the staff of the U B C mass spectrometry facility using a Kratos MS-50 spectrometer. Elemental analyses were performed by Mr. Minaz Lakha of the U B C microanalytical facility. Dr. Peter Graham prepared compounds 8-15 and 19-23. Complete characterization data and key details of synthesis are included here for purposes of comparison to the related tungsten products (Chapter 4) and of mechanistic insight (Chapter 5).  3.4.2 Preparation of 7 A red solution of 3 (0.079 g, 0.196 mmol) in cyclopentene (10 mL) became orange after 26 h. The final solution was then evaporated to dryness in vacuo, and the residue was recrystallized from pentane/Et20 at -30 °C ovemight to obtain 7 as an orange powder (0.039 g, 0.135 mmol, 69%).  Anal. Caled for C20H33M0NO: C, 60.14; H , 3.51; N , 8.33. Found: C, 60.13; H , 3.90; N , 8.06; IR (Nujol): V N O = 1559 cm-'; ' H N M R (CeDe, Mo-  400 MHz, 25 °C) 6 7.94 (IH, dd, V H H = 18.7, ^ J H H = 8.3, H2), 3.27 ( I H ,  O  d,  1  V H H  = 5.2, H6), 2.42 (IH, ddd, V H H = 12.4, V H H = 8.3,  VHH  = 5.6,  H3), 1.94 (IH, m (overlapping), H4), 1.88 ( I H , m (overlapping), H5), 1.80 (IH, m (overlapping), H3), 1.74 (15H, s, CjMes), 1.70 (IH, m (overlapping), H4), 1.15 (9H, s, CMe3), 0.08 ( I H , ddd, V H H = 23.6, V H H = 12.4, V H H = 5.2, H5), -0.05 (IH, ddd, V H H = 18.7, VHH  =10.1,  V H H  = 5.2, Hl);  '^CÍ'H}  N M R (CeDe, 100 MHz, 25 °C) 5 154.2 (C2), 122.7 (C6),  109.7 (CsMes), 41.6 (CMe3), 39.8 (C4 or C5), 38.4 (C3), 35.9 (C4 or C5), 32.7 (CMe3), 20.4 (Cl), 10.7 (CsMes). M S (El, 120 °C): miz 401 [P^].  3.4.3 Preparation of 8 A THF solution of 7 was prepared and pyridine was added. The orange solution tumed brown over 18 h. Product 8 was obtained as orange crystals in 35% yield by chromatography and recrystallization from Et20 at -30 °C. ce  5:  M o^H  IR (Nujol): V N O = 1609 cm"'; ' H N M R (CeDe, 400 MHz, 25 °C) 6 2.74 ( I H , dd, V H H  V H H  = 13.3, V H H = 7.6, H3), 2.63 (IH, ddd, V H H = 13.3, V H H = 10.5,  = 3.5, H3), 2.54 (IH, dd, V H H = 14.5, V H H = 8.1, H5\ 2.24 (IH, ddd.  V H H = 14.5, V H H = 11.1, V H H = 8.1, H5), 1.86 (IH, m (overlapping), H4), 1.76 ( I H , m (overlapping), H4), 1.73 (15H, s, CjMes), 1.61 (IH, br s, H2), 1.59 (IH, br s, H6), 1.38 (9H, s, CMe3), -2.04 (IH, s, Mo-H); '^Cf'H} N M R (CeDe, 100 MHz, 25 °C) 6 123.2 ( C l ) , 106.1 (CjMej), 81.8 (C2), 80.0 (C6), 36.4 (CMe3), 35.4 (C3), 33.8 {CMe^), 33.2 (C5), 24.2 (C4), 11.7 {CsMes). HRMS-EI miz: [M]* caled for ^^MoC2oH33NO 401.16162; found 401.16080.  3.4.4 Preparation of 9 A red solution of 3 in cyclooctene diluted with cyclohexane darkened to brown añer 18 h. Product 9 was obtained as red crystals in 51% yield by chromatography and recrystallization from pentane at -30 °C. Anal. Caled for C23H39M0NO: C, 62.57; H , 8.90; N , 3.17. Found: C, 62.81; H , 8.60; N , 3.46; IR (Nujol): V N O = 1586 cm"'; ' H N M R (CeDe, 400 MHz, 25 °C) 5 4.31 (IH, d, V H H = 4.4,  H9), 4.10 (IH, ddd, 3  (IH, ddd,  4  Mo--4O  1  V H H  2.19 (IH, m,  V H H  = 11.7, V H H = 4.0, V H H = 3.4, H2\ 2.79  = 14.3, V H H = 5.8,  V H H  = 3.0, H3), 2.45 ( I H , m, H3),  1.82 (IH, m (overlapping), H4), 1.73 ( I H , m  (overlapping), H5), 1.69 (15H, s, CsMej), 1.64 (IH, m (overlapping), 8  7  H7), 1.51 (IH, m (overlapping), H4), 1.50 (IH, m (overlapping),  H5), 1.48 (IH, m (overlapping), H6), 1.46 (IH, m (overlapping), H7), 1.36 (IH, m (overlapping), H6), 1.28 (9H, s, CMes), 0.20 (IH, m, H8\ -0.18 (IH, ddd, V H H = 11.7, V H H = 7.5, V H H = 4.4, Hl);  ' ^ C { ' H } N M R (CeDe, 100 M H z , 25 °C) 8 152.3 (C9), 121.5 (C2), 109.3  (CsMes), 44.4 (C3), 44.1 (CMea), 36.1 (C8), 32.9 (CAfeg), 30.5 (C4), 28.8 (C7), 28.2 (C5 or C6), 28.1 (C5 or C6), 14.9 ( C l ) , 10.7 (CsMes); M S (El, 100 °C): m/z 441 [P-2"'] (H2 is lost upon heating).  3.4.5 Preparation of 10 A n orange solution of 9 in cyclohexane was heated at 50 °C for 48 h. Following chromatography, the yellow eluate was evaporated in vacuo to obtain 10 as a yellow solid in 50% yield. Anal. Caled for C23H37M0NO: C, 62.86; H , 8.49; N , 3.19; Found: C,  Mo o  v  3 4  62.45; H , 8.59; N , 3.38; IR (Nujol): 400 M H z , 25 °C) 5 3.06 (IH, ddd,  VNO=  V H H  1595 cm''; ' H N M R (CéDg,  = 16.3, V H H = 10.3, V H H  =  8.0, H3), 2.83 (IH, ddd, V H H =17.1, V H H = 5.9, V H H = 3.9, H8), 2.33 8  7  (2H, m (overlapping), H4 and H6), 2.21 (IH, m (overlapping), H8), 2.14 (IH, m (overlapping), H4), 1.89 (IH, m, H7), 1.81 (IH, m  (overlapping), H7), 1.75 (IH, m (overlapping), H5), 1.67 (15H, s, CjMes), 1.66 (IH, s, H9), 1.57 (IH, m, H5), 1.40 (IH, m, H6), 1.30 (9H, s, CMea), 0.72 (IH, d, V H H = 8.0, H2). ' ^ C { ' H } N M R  (CóDé, 100 MHz, 25 °C) 5 121.3 ( C l ) , 105.6 (CjMes), 96.7 (C9), 85.2 (C2), 78.0 (C3), 34.9 (CMea), 33.3 (CM^s), 31.8 (C5), 31.3 (C4), 30.1 (C7), 29.0 (C8), 28.0 (C6), 11.3 (CsMes); M S (El, 100 °C): m/z 441 [P^].  3.4.6 Preparation of 11 A yellow solution of 10 in neat pyridine was heated at 100 °C for 19 h. Chromatography on an alumina column separated the remaining starting material from the red-orange product.  The red residue was recrystaUized at -30 °C overnight to obtain 11 as an orange precipítate in 72% yield. Anal. Caled for C28H42M0N2O: C, 64.85; H , 8.16; N , 5.40; Found: C,  3  64.84; H , 8.49; N , 5.80; IR (Nujol): V N O = 1552 cm-'; ' H N M R (C^De,  4  400 M H z , 25 °C) 5 8.07 (2H, br s, ortho py), 6.64 (IH, t, V H H = 7.5, O  •N, /  1  \ 8  6  para py), 6.39 (2H, t, V H H = 6.4, meta py), 4.80 (IH, s, H9), 3.13 (IH, m, H3), 2.81 (IH, m, H7), 2.57 (IH, m, H4), 2.45 ( I H , dd, V H H  9  =  13.7, V H H = 8.6, H7), 2.27 (IH, m, H3), 1.93 (2H, m (buried), both  H6), 1.90 (IH, buried, Hl), 1.87 (IH, buried, H2), 1.75 (IH, m, H4), 1.64 ( I H , m, H5), 1.53 (15H, s, CsMes), 1.37 (IH, m, H5), 0.95 (9H, s, CMes); ' ^ C I ' H } N M R (CeDe, 100 MHz, 25 °C) S 154.6 (ortho py), 145.1 (C8 (observed by HMBC)), 135.0 (parapy), 129.3 (C9), 124.6 (meta py), 105.4 (CsMes), 77.1 (Cl), 61.9 (C2), 34.7 (C6), 33.1 (CMe^), 32.8 (C5), 32.2 (C3), 28.0 (CMes), 26.2 (C7), 23.1 (C4), 9.7 (CsMes); M S (El, 120 °C): m/z 520 [P^].  3.4.7 Preparation of 12 A yellow solution of 10 in cyclopentene was heated at 70 °C for 3 d. The reaction mixture was evaporated to dryness in vacuo, extracted with pentane and chromatographed on an alumina column to obtain 3-neopentylidene-cyclooctene (12) as a colorless oil (47% yield). ' H N M R (CóDó, 400 MHz, 25 °C) 5 6.21 (IH, d, V H H = 12.0, H2), 5.57 (IH, s, H9), 5.37 (IH, ddd, V H H = 12.0,  V H H  = 8.6, V H H = 3.4, Hl), 2.64 (2H, t,  V H H  =  6.5, both H4), 2.33 (2H, dd, V H H = 14.9, V H H = 8.6, both H8), 1.62 (4H, m, both H5 and both H6), 1.48 (2H, m, both H7), 1.09 (9H, s, CMes); '^Cf'H} N M R (CéDe, 100 M H z , 25 °C) 5 144.5 (C9), 140.3 (C2), 138.6 (C3), 124.8 (Cl), 32.0 (CMe2), 28.6 (C7), 28.2 (CMes), 28.1 (C4), 26.5 (C8), 23.2 (C5 or C6), 23.1 (C5 or C6); M S (El, 100 °C): m/z 178 [ P ^ ] .  3.4.8 Preparation of 13 A red solution of 3 in mixed cycloheptene and cyclohexane darkened to red brown añer 6 h. The solution was then evaporated to dryness in vacuo, and the residue was dissolved in CóDe for spectroscopic characterization as a mixture of 13 and 14.  IR (Nujol): V N O = 1588 cm-'; ' H N M R (CeDg, 400 M H z , 25 °C, 3  A  selected resonances) 6 5.66 (IH, dd, V H H = 12.7, V H H = 9.9, H2), 5  4.52 (IH, d, V H H = 4.2, H8), 2.20 (IH, br d, V H H = 12.7, H3), 2.00 (IH, m (buried), H7), 1.72 (15H, s, CsMes), 1.60 (IH, m (buried), H3), 1.22 (9H, s, CMes), 0.41 (IH, dd, V H H = 23.0, V H H = 11.0, H7),  -0.46 (IH, dd, V H H = 9.9, V H H = 4.2, Hl); ' ^ C { ' H } N M R (CéDe, 100 M H z , 25 °C, selected resonances) 5 155.5 (C8), 110.2 (C2), 109.6 (CjMes), 36.8 (C3), 32.5 {CMe^}, 31.9 (C7), 10.9 {CsMes).  3.4.9 Preparation of 14 A red solution of 3 in mixed cycloheptene and cyclohexane tumed brown after 20 h. The final solution was evaporated to dryness in vacuo, the resulting residue was dissolved in pentane, and the solvent was allowed to evapórate over 2 days to give dark red crystals of 14 in 25%  IR (Nujol): V N O = 1586 cm''; ' H N M R (CeDg, 400 MHz, 25 °C) 5 4.34 (IH, ddd, V H H = 9.2,  V H H  = 5.3,  V H H  = 5.3, H2), 3.71 (IH, d,  V H H = 6.1, //5), 2.70 (2H, m, both H3), 2.35 (IH, dddd, V R H = 13.3, V H H = 6.3, V H H = 6.3, V H H = 2.4, H7), 1.70 (15H, s, CsMes), 1.70 (IH, m (under CsMes), H4), 1.60 (2H, m (overlapping), H4 and //(5), 1.58 (IH, m (overlapping), H6), 1.54 ( I H , m (overlapping), H5), 1.43 (IH, m (overlapping), H5), 1.27 (9H, s, CMej), 0.35 (IH, ddd, V H H = 20.0, V H H = 13.3, V H H = 6.3, H7), -0.28 ( I H , dddd, V H H = 20.0, V H H = 9.2, V H H = 6.1, V H H = 2.4, Hl); ' ' C { ' H } N M R  (CóDfi, 100 M H z , 25 °C) 5 138.2 (C8), 130.9 (C2), 109.3 (CsMes), 41.6 (C3), 32.9 {CMe^), 32.7 (C4), 31.5 (C6), 30.0 (C7), 25.2 (C5), 15.2 (Cl), 10.7 (CsMes). M S (El, 120 °C): miz All [P-2*].  3.4.10 Preparation of 15 A red solution of 3 in mixed cycloheptene and cyclohexane was heated at 50 °C for 20 h. The reaction residue was chromatographed to obtain a yellow eluate that was evaporated in vacuo to obtain 15 as a yellow solid in 35%) yield.  IR (Nujol): V N O = 1589 cm-'; ' H N M R (CeDe, 400 MHz, 25 ° C ) 5 3.21 (IH, ddd, V H H = 9.0,  VHH  = 4.5, V H H = 3.0, H3), 2.90 (IH, d, V H H  17.1, H7), 2.72 (IH, m, H4), 2.61 ( I H , m, H7), 2.42 (IH, d, V H H  o  =  =  15.2, H4), 1.68 (IH, m (overlapping), H5 or H6), 1.67 (IH, m (overlapping), H5 or H6), 1.66 (IH, m (overlapping), H5 or //(5), 1.64 (15H, s, CsMes), 1.63 (IH, m (overlapping), H5 or H6), 1.40 ( I H , s,  H8), 1.33 (9H, s, CMea), 0.21 (IH, d, V H H = 9.0, 7/2); ' ^ C Í ' H } N M R (CeDe, 100 MHz, 25 °C) 6 123.2 (Cl), 105.5 (CsMes), 93.6 (C8), 82.2 (C3), 81.1 (C2), 35.5 (CMes), 33.3 (C7), 33.0 (CMei), 33.0 (C4), 26.5 (C5 or C6), 26.4 (C5 or C6), 10.9 (CsMes). M S (El, 100 °C): m/z 427  3.4.11 Preparation of 16 A red solution of 3 (0.047 g, 0.116 mmol) in cyclohexene (6.64 g) became brown after 110 h. The solution was then evaporated to dryness in vacuo, and the residue was extracted with pentane ( 2 x 2 mL). The pentane extracts were then loaded onto the top of a column of neutral alumina (ca. 2.5 x 0.5 cm) prepared with pentane. The column was washed with pentane (15 mL). A yellow band was then eluted with 9:1 pentane/diethyl ether (15 mL) and coUected. The yellow eluate was evaporated to obtain 16 as a yellow solid (0.018 g, 0.044 mmol, 38%). X-ray quality crystals of 16 were obtained by dissolving the compound in diethyl ether and slowly allowing the solvent to evapórate over a period of days at room temperature. IR (Nujol): V N O = 1598 cm-'; ' H N M R (CeDe, 500 MHz, 25 °C) 6 3.74 (IH, ddd, V H H = 6.7, V H H = 6.7, „.Mo 2 > ^ ^ 4 7  VHH  = 1.0, H3), 2.52 (IH, dd, V H H  =  16.5, V H H = 4.9, H5), 2.30-2.40 (3H, m (overlapping), both H4 and H5), 1.70 (IH, s,H7), 1.65 (15H, s, CsMes), 1.54 (2H,m, hoÜiHÓ), 1.32 (9H, s, CMe3), 1.12 (IH, d, V H H = 6.3,7/2); ' ^ C Í ' H } N M R (CeDe, 125 MHz, 25 °C) 8 115.8 ( C l ) , 105.2 (CsMes), 94.1 (C7), 85.9 (C2), 76.9 (C3),  34.9 (CMes), 32.9 (CMe^), 27.9 (C5), 26.8 (C4), 22.8 (C6), 10.8 (CsMes); M S (El, 120 °C): m/z 413 [P*]. A second (blue) band eluted añer the yellow band. The blue-green residue was highly soluble in pentane. The blue compound was impossible to isolate in a puré form. ' h  NMR  spectra always contained large signáis arising from 16. Redissolving the blue product in  cyclohexene at room temperature resulted in a color change from blue to yellow-brown añer 5 days. ' H N M R spectroscopy identified the main product of this reaction as 16. The blue compound was not identified.  3.4.12 Experimental evidence for 17 A sample of 3 (64 mg) was dissolved in cyclohexene (15.0 mL) and allowed to react at room temperature. At measured time intervals, 1 mL aliquots were removed with a marked pipet, placed in labeled vials, and immediately stripped of the volátiles in vacuo to isolate the organometallic species. On the first day, the first aliquot was removed after 20 minutes reaction time and subsequent aliquots in 2-hour intervals thereafter. Specifically, aliquots were removed at 20 min (solution was red), 140 min, 260 min, 380 min (solution was orange-red), 500 min, 620 min (solution was médium orange), 735 min, 1460 min (2"^* day, solution was yellowbrown), 1580 min, 1700 min, 1820 min, 1940 min (solution was still yellow-brown), 2880 min (48 h, 3'''' day), and finally 5820 min (97 h, 4* day). The residue of each aliquot was dissolved in CeDe (1 mL, measured by syringe), and analyzed by ' H N M R spectroscopy. Comparisons of the time-lapsed spectra showed that the peaks due to 3 decrease and disappear by 1460 min. Peaks for a new compound (17) appear by 260 min and then decrease again. Peaks arising from 16 appear very small at 260 min, grow steadily, and remain as the only major peaks due to an organometallic product at the end of the reaction time (5820 min). Figure 3.10 shows a qualitative attempt to portray these trends. In the N M R spectra some traces of decomposition appeared in addition to the major compounds of interest. Also, small fraces of organic products appeared throughout the reaction sequence. Selected peaks attributed to 17, ' H N M R (CóDe, 400 M H z , 25 °C) 5 3.91 (IH, d, H7), 1.74 (15H, s, CjMes) 1.22 (9H, s, CMes) 0.1 (IH, ddd oxm, H2), -0.7 (IH, m,Hl).  o  10  20  30  40  50  60  70  80  90  100  Reaction time (h)  Figure 3.10 The relative percentages of compounds 3,17 and 16 present in the reaction mixture are determined by dividing the integrated valué of each compound's Cp* peak by the sum of the integrated valúes for all three Cp* peaks, for each aliquot withdrawn over the reaction time. Trend lines are drawn as a guide to the eye only, and error bars are set at 10%.  3.4.13 UV/vis experiment investigating the reaction of 3 with cyclohexene The reaction of 3 with cyclohexene was monitored by UV/vis spectroscopy. A narrow 0.1 cm path-length resealable quartz cell equipped with a large side-arm reservoir was used. First, the necessary concentrations and the characteristic spectra of 3 and 16 were determined. Concentrations of ca. 1 x 10"^ M gave appropriate absorbance valúes of less than 2. The characteristic UV/vis spectrum of 3 in cyclohexene had three main peaks, at 228 nm (e = 5.8 X 10^), 292 nm (£ = 3.2 X 10^) and 482 nm (s = 2.2 x 10^). The characteristic spectrum of an authentic sample of 16 in cyclohexene showed one sepárate peak at 230 nm (s = 1.0 x lO'') and a second peak at 286 nm (s = 5.3 x 10^) with a strong shoulder at about 330 nm (s = 3.0 x lO'). For the timed trial, 3 (0.016 g) and cyclohexene (30.0 mL) were placed in the quartz cell, initially separated. They were mixed thoroughly to form a red solution (0.0013 M), and the cell was placed in the holder. The temperature of the holder and the cell was controlled at 25 °C. Scans were taken at the following times (min) after mixing: 2, 3, 13, 23, 33, 48, 63, 78, 93, 108,  123, 138, 153, 168, 183, 198, 213, 228, 243, 258, 273, 288, 347, 407, 469, 503, 572, 1442*, 1587, 1708, 1821, 2810, 2958, 3273, 4333, and 4533. (* Note that this time and all subsequent times were after an inadvertent interruption to the lamp. As a result a second blank was used to recalibrate the machine and thus the scans at and after 1442 minutes do not correspond perfectly to the former ones.) ' H N M R analysis of the final products after completion of the UV/vis analysis showed 16 was indeed the main product. The wavelength and absorbance data of the diagnostic peaks were compiled and analyzed using Excel spreadsheets. For illustrations, see Figures 3.8 and 3.9.  3.4.14 Spectroscopic and GC/MS evidence for organic products of cyclohexene ' H N M R spectra of the crude reaction mixtures resulting from 3 in cyclohexene often displayed small signáis at ~ 5.7 ppm, which corresponded to olefinic protons and looked much like the similar signáis obtained for authentic cyclohexene oligomers isolated from the reaction of cyclohexene and Cp*W(NO)(CH2CMe3)2 (see Chapter 2). One such ' H N M R spectrum is shown below in Figure 3.11 as an example. Further reactions of 16 with cyclohexene also showed similar traces in the ' H N M R data (vide infra).  1. "T '  " 1 6  , '  T 5  '  1 4  _ ' • -I— - ' 3  > Í n S  1  "r• 1  '  1  Figure 3.11 ' H N M R spectrum (in C(,D(,) of the crude product mixture obtained from the reaction of 3 with cyclohexene at rt for 3 days. Peaks at 1.6 ppm and 1.3 ppm are due to 16, and peaks at 5.7 ppm are due to the olefinic protons of the oligomeric products.  The oUgomeric organic products were isolated from the reaction mixtures by column chromatography on alumina with either pentane or hexanes as the mobile phase. The isolated organic products were dissolved in Et20, and analyzed by the standard GC/MS method. The resulting GC trace showed peaks at 1.4 min (m/z = 164, 152, 169; dimers of cyclohexene with unsaturations), 1.9 min (m/z = 158, 154; neopentylcyclohexene), and múltiple small peaks in the 9 to 11 min región (m/z = 230; neopentylcyclohexenylcyclohexene. m/z = 246, 244, 242; trimers of cyclohexene with unsaturations.) Figure 3.12 shows a typical G C trace for the oligomers obtained from a reaction of 3 with cyclohexene.  ———  ——  _____  -nCfHBrSTD  •  —'  '  '  iBoom 260CXX» 2400000 2200000  2(mm 1800000 1600000 1400000 1200000 1000000 socnoo 600000 400000 200000 nme->  iio  3Í»  4ÍM ¿ i »  é.bb  7.1»  ¿.i»  a b ó m ' ó o ii.'dd i z ' o ó  is'áo ulóo ÍSM  iéi»  W.áá i k o o  lolod 2b:ofj a i l o d z i o o  Figure 3.12 G C trace of isolated oligomeric products from the reaction of 3 with cyclohexene (standard method), showing a peak at 1.4 min (dimers), at 1.9 min (neopentylcyclohexene), and múltiple small peaks in the 9 to 11 min región (neopentylcyclohexenylcyclohexene, trimers of cyclohexene).  3.4.15 Reactivity of 16 with cyclohexene under thermolysis conditions Thermolysis experiments of 16 in cyclohexene were performed at various temperatures (rt, 50 °C, 70 °C, and 100 °C) in order to explore the further reactivity of 16 with regard to the formation of oligomers of cyclohexene. These experiments are presented below. Table 3.2 summarizes the key comparisons.  Room temperature: In the glovebox, yellow 16 (10 mg) was dissolved in cyclohexene (10.0 mL) and left to sit undisturbed for 11 days, at which point the solution was golden-yellow (i.e. only slightly darker than when the reaction started). Volátiles were removed in vacuo. ' H N M R analysis of the crude reaction mixture (in CeDe) showed the signáis for 16 and one small additional peak at 5.7 ppm, attributed to the olefmic protons of the organic products (oligomers of cyclohexene). The relative integration ratio of the Cp* peak due to 16 and the signáis due to the oligomeric olefínic protons was 32 to 1. 50 °C: Compound 16 (5 mg) was dissolved in cyclohexene (7 mL), placed in a resealable reaction vessel, and heated with an oil bath at 50 °C for 9 days. The final solution was yellowbrown, and the residue remaining after the removal of volátiles in vacuo was brown. ' H N M R analysis of the crude reaction mixture showed 16 as the only organometallic compound present. Peaks between 5.5 and 6 ppm showed the formation of the organic products. The relative integration ratio of the Cp* peak due to 16 and the signáis due to the oligomeric olefmic protons was 2 to 1. 70 °C: Coumpound 16 (24 mg) was dissolved in cyclohexene (6.0 mL), placed in a resealable reaction vessel, and heated with an oil bath at 70 °C for 3 weeks. The final solution was dark red-brown. The residue after removal of the volátiles was dark, and 94 mg of products were recovered. ' H N M R analysis of the crude products showed the presence of 16 and organic products (Figure 3.13). The relative integration ratio of the Cp* peak due to 16 and the signáis due to the oligomeric olefinic protons was 1 to 10. The products were separated on an alumina column (ca. 2.5 cm x 0.5 cm). A n initial pentane fraction contained the organic products (100 mg), and a subsequent pentane/Et20 fraction recovered the unreacted 16 (23 mg). The brown decomposition products remained on the column. Analysis of the organic fraction by GC/MS showed the presence of cyclohexene dimers and trimers, both with relatively high levéis of unsaturation (2 or 3 per oligomer on average, which is more than the 1 or 2 unsaturations per oligomer typical of the timgsten system). The analysis also showed the presence of compounds having a neopentyl group coupled to a cyclohexene dimer (or trimer) molecule. 100 °C: Compound 16 (6 mg) was dissolved in cyclohexene (approx. 8 mL), placed in a resealable reaction vessel, and heated with an oil bath at 100 °C for 2 days. The solution tumed dark brown. The volátiles were removed in vacuo. ' H N M R analysis of the cmde products showed the expected organic products, and the complete loss of 16. No integration ratio could be determined.  I • i ' " ( •' I - r - I ' r l § 5 í 3 2 J Figure 3.13 H N M R spectram (in CgDe) of the crade product mixture obtained from the P£5  reaction of 16 in cyclohexene at 70 °C for 3 weeks. Peaks at 1.6 ppm and 1.3 ppm are due to remaining 16, and peaks at 5.5-6.0 ppm are due to the olefmic protons of the oligomeric products.  Table 3.2 Summary of the relative ratio of remaining 16 and oligomeric organic products at various reaction temperatures and times, based on the ' H N M R integration valúes of the Cp* peak due to 16 and the olefinic protón peaks due to the organic products. Temp.  Time  Amounts  Ratio  Crude yield  mg of 16  mL cyclohexene  Cp* : olefín H  (mg, all products)  rt  lid  10  10.0  32 : 1  -  50 °C  9d  5  7.0  2 : 1  -  70 °C  3 wks  24  6.0  1 : 10  94  100 ° c  id''  10  10.0  1:4  49  100 ° c  2d  6  8.0  See section 4.4.16 below for reaction details.  undeteiinined  -  A ratio can not be determined due to total consumption of 16.  3.4.16 Further thermal reactivity of 16 in cyclohexene and in mixed solvents Comparisons of the thermolysis of 16 in neat cyclohexene and in mixed cyclohexene/cyclohexane were performed in the following manner. In the first reaction, 16 (10 mg) was dissolved in neat cyclohexene (10.0 mL) and placed in a resealable reaction vessel. In the second reaction, 16 (14 mg) was dissolved in cyclohexene (1.0 mL) and cyclohexane (9.0 mL) and placed in a second resealable reaction vessel. Both reaction vessels were heated at 100 °C for 25 hours. Both solutions initially appeared light yellow, and both changed to yellow-brown with a dark powdery precipítate that settled on the bottom of the flasks. The volátiles were removed in vacuo for both reaction mixtures. The first reaction gave a residue that was dark and oily, while the second gave a residue that was also dark but not oily. The recovered crude yields were 49 mg and 19 mg, respectively. ' H N M R analyses of the product residues (in CeDg) showed marked differences. The spectrum for the first reaction gave signáis for remaining 16 and for oligomeric organic products. (This spectrum looked very similar to the one shown above in Figure 3.13.) The spectrum for the second reaction gave only signáis for remaining 16, with no signs of any oligomeric products. Although there were signs of decomposition, the ' H N M R spectroscopic data gave no clues as to the nature of the decomposition products.  3.4.17 Preparation of 18 CpMo(NO)(CH2CMe3)2 was made according to the previously published procedure,'' starting with CpMo(NO)Cl2 (0.143 g). The products were filtered and isolated as a bright red residue. Then cold cyclohexene (ca. 10 mL) was vacuum transferred onto the residue, and as it thawed and warmed to rt it produced a red solution. Half the solution was cannulated into a second Schlenk flask and placed in the dark. The other Schlenk flask remained on the benchtop in the light. Both reaction mixtures were worked up in a similar manner after one day, and both yielded the same product. After one day at rt the solutions had tumed brown and contained dark precipítales. The cyclohexene solvent was removed in vacuo, and the resulting residue was extracted with pentane to obtain a yellow solution. After ' H N M R analysis showed the organometallic product to be the same, the pentane extracts were combined, and recrystallization was attempted from hexanes/Et20. This yielded a yellow powdery solid 18 which was analyzed by conventional methods. (No oligomeric products were detected in the reaction mixture.)  C16H33M0NO IR (Nujol): V N O = 1616 cm''; ' H N M R (CeDe, 500 M H z ,  25 °C) 6 4.97 (5H, s, C5H5), 4.41 (IH, dt, V H H = 6.3, V H H = 3.5, H3), •^•"^°2p^^  2.28-2.39 (4H, m, both//^ and bothi/d), 2.10 (IH, s, H7), 1.94 ( I H , d.  •  V H H = 6.3, H2), 1.45-1.54 (2H, m, both H5), 1.22 (9H, s, CMes); '^C{*H} N M R (CeDe, 125 MHz, 25 °C) 6 95.1  (C5H5), 88.2 (C7), 78.0  (C2), 70.0 (C3), 35.1 (CMea), 32.5 (CMej), 26.5 (C4 or C6), 26.2 (C4 or C6), 22.5 (C5), C l quat obscured; M S (El, 100 °C): m/z 343 [P*].  3.4.18 Preparation of 19 Compound 3 was dissolved in 1-hexene and diluted with cyclohexane. The resulting pink solution darkened to a brovm color over 18 h. Chromatography of the reaction residue produced a yellow eluate that was evaporated to obtain 19 as a yellow solid in 29% yield. IR (Nujol): V N O = 1600 cm-'; ' H N M R (CeDg, 400 MHz, 25 °C) 6 3.36 ( I H , ddd, V H H = 13.7, V H H = 10.7, V H H = 3.5, H3), 3.18 (IH, dd,  _  VHH  = 12.6, V H H = 10.5, 7/7), 2.11 (IH, d, V H H = 12.6, 7/7), 1.97 ( l H , m , H4), 1.68 (15H, s, CsMes), 1.59-1.53 (4H, m (overlapping), 772, H4  _ / 6  3  and both 775), 1.24 (9H, s, CMca), 1.02 (3H, t, V H H = 7.2, all H6);  5  13  C { ' H } N M R (CeDe, 100 MHz, 25 °C) 5 105.4 (CsMes), 99.6 (C7),  89.3 ( C l or C3), 88.7 ( C l or C3), 77.6 (C2), 35.6 (C4 or C5), 35.5 (CMea), 33.6 (CMej), 28.3 (C4 or C5), 14.3 (C6), 11.3 (CsMes); HRMS-EI m/z: [M]^ caled for ^^MoC2iH35NO 415.17727; found 415.17742.  3.4.19 Preparationof20 Complex 20 was prepared in a manner analogous to 19 using 1,7-octadiene (42 h, 33%). IR (Nujol): V N O = 1600 cm-'; ' H N M R (CeDe, 400 MHz, 25 °C) 6 5.83 (IH, ddt, V H H = 16.8, V H H = 10.2, V H H = 6.7,777), 5.08  ,Mo^ O  / _ / 2 6  8 7  y 5  3  9  1  (IH, ddd, V H H = 16.8, V H H = 3.3, VHH=  V H H  = 1-5, H8), 5.03 (IH, d,  10.2, 775), 3.31 (IH, ddd, V H H =13.1,  V H H  = 11-2, V H H  =  3.9, H2), 3.18 (IH, dd, V H H = 13.1, V H H = 10.2,775), 2.16 (2H, overlapping, both 77(5), 2.14 (2H, overlapping, 77^), 2.13 (IH, overlapping, 779), 2.10 (2H, overlapping, both 775), 1.68 (15H, s.  CsMes), 1.48 (IH, m, Hl), 1.24 (9H, s, CMej); " C { ' H } N M R (CóDg, 100 M H z , 25 °C) 6 139.4 (C7), 115.2 (C8), 105.5 (CsMes), 99.6 (C9), 89.4 (C3), 88.6 (Cl), 77.3 (C2), 35.5 (CMes), 34.4 (C4, C5, or C6), 34.0 (C4, C5, or C6), 33.6 {CMe^), 32.8 (C4, C5, or C6), 11.3 (CsMes); M S (El, 100 °C): m/z 441 [P^].  3.4.20 Preparation of 21 and characterization of 23 Compound 3 was dissolved in allylbenzene and diluted with cyclohexane. The resulting pink solution darkened to a brown color over 18 h. The reaction residue was dried in vacuo and chromatographed on an alumina column. The pentane fraction yielded a colorless eluate from which 23 (5.6 eq) as obtained as a colorless liquid (see characterization data below). A 4:1 pentane/diethyl ether fraction yielded a yellow eluate that was evaporated to dryness in vacuo; the residue was recrystallized from pentane at -30 °C to obtain 21 as yellow crystals (24% yield). IR (Nujol): V N O = 1603 cm''; ' H N M R (CeDg, 400 MHz, 25 °C) 6 6.957.27 (5H, m (overlapping), aryl CH), 4.46 (IH, d, V H H = 13.3, H3), 10  _  3.43 (IH, dd,  V H H  = 12.0,  V H H  = 10.0, Hl), 2.43 (IH, dd,  V H H  = 13.3,  V H H = 10.0, H2), 2.25 (IH, d, V H H = 12.0, HIO), 1.46 (15H, s, CsMes),  O  Ph-  1.24 (9H, s, CMes); ' ^ C { ' H } N M R {C¿De, 100 M H z , 25 °C) 6 140.9  3  (aryl quat C), 128.9 (aryl C), 128.8 (aryl C), 128.7 (aryl C), 126.0 (aryl C), 106.1 (CsMes), 99.3 (CIO), 89.3 ( C l or C3), 82.2 ( C l or C3), 76.6 (C2), 35.6 (CMes), 33.5 (CMes), 10.6 (CjMes); HRMS-EI miz: [Uf caled for ^^MoC24H33NO 449.16162; found 449.16166. (£)-(4-methylpent-l-ene-l,5-diyl)dibenzene (23): ' H N M R (CeDg, 400 MHz, 25 °C) 6 6.68-6.45 (lOH, m overlapping, aryl CH), 5.72 (IH, d, V H H VHH  14.0,  6.8,  = 13.3,  V H H  V H H = 7.2, V H H  = 15.9, Hl), 5.50 (IH, dt,  = 6.3, H5), 1.70 (IH, dd,  H3),  1.33  (IH,  V H H  dt, V H H = 14.0,  V H H  = 13.3,  = 15.9,  V R H  V H H = 7.2,  V H H  = 7.2, H2), 1.98 (IH, dd,  = 6.3, H5), 1.55 (IH, dt,  H3),  1.18  (IH,  ddd,  V H H =  V H H = 7.2,  V H H =  = 6.3, H4), 0.25 (3H, d, V H H = 6.8, Me); ' ^ C { ' H } N M R (CéDg, 100 MHz, 25 °C) 5  141.8 (aryl quat C), 138.6 (aryl quat C), 132.3 (Cl), 129.9 (2 aryl C), 129.6 (C2), 129.1 (2 aryl C), 128.8 (2 aryl C), 127.5 (aryl para C), 126.8 (2 aryl C), 126.5 (aryl para C), 43.9 (C5), 40.75 (C3), 36.1 (C4), 19.9 (Me); M S (El, 150 °C): miz 236.1 [P*].  3.4.21 X-ray Crystallography Data coUection for each compound was carried out at -100 ±1 °C on a Rigaku AFC7/ADSC C C D diffractometer, using graphite-monochromated Mo K a radiation. Data for 7 were coUected to a máximum 26 valué of 55.2 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 3600 observed reflections and 216 variable parameters. Data for 8 were coUected to a máximum 26 valué of 55.8 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms HOl (the hydride), H2, and H6 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 4608 observed reflections and 228 variable parameters. Data for 9 were coUected to a máximum 26 valué of 49.8 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l , H2, and H9 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 3966 observed reflections and 255 variable parameters. Data for 10 were coUected to a máximum 26 valué of 53.1 ° in 0.5 ° oscUlations. The structure was solved by direct methods^ and expanded using Fourier techniques. The cyclooctenyl ring was disordered at C5, C6 and C7. The disorder was modeled with two orientations of 40% and 60% occupancy. Constraints were used to keep the related bond lengths in the disordered región equivalent. The three disordered carbons were refined isotropically; all other non-hydrogen atoms were refined anisotropically. Hydrogen atoms H2, H3 and H9 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 4201 observed reflections and 255 variable parameters. Data for 11 were coUected to a máximum 26 valué of 55.1 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. A l l non-  hydrogen atoms were refined anisotropically; hydrogen atoms H l , H2, and H9 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 5599 observed reflections and 309 variable parameters. Data for 16 were coUected to a máximum 20 valué of 55.0 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. The coupled organic ligand and the nitrosyl ligand were disordered in two orientations. Restraints were used to maintain similar geometries for the related segments, namely the tert-buty\ groups, the cyclohexane rings, and the nitrosyl ligands. Overlapping atoms (NI and C23, N2 and C2, C8 and C29, and C l 1 and C32) were refined isotropically; all other non-hydrogen atoms were refined anisotropically. Hydrogen atoms H2, H3, H23, and H24 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of ñiU-matrix least-squares analysis was based on 4157 observed reflections and 322 variable parameters. For each structure neutral-atom scattering factors were taken from Cromer and Waber.^ Anomalous dispersión effects were included in  Fcaic;^  the valúes for Af and A / ' were those of  Creagh and McAuley.^ The valúes for mass attenuation coefficients are those of Creagh and Hubbell.^ A l l calculations were performed using the CrystalClear software package of Rigaku/MSC,'*^ or Shelxl-97.'' X-ray crystallographic data for the six structures are presented in Table 3.3, and in the cif files.  Table 3.3 X-ray Crystallographic Data for Complexes 7, 8 and 9. 7  8  9  Empirical formula  C20H33M0NO  C20H33M0NO  C23H39M0NO  Crystal Habit, color  Prism, orange  Píate, yellow  Irregular, red-orange  Crystal size (mm)  0.5x0.3x0.1  0.6x0.4x0.1  0.25  Crystal system  Triclinic  Monoclinic  Monoclinic  Space group  P-1  P2i/c  C2/C  Volume (Á^)  945.7(2)  1939.8(1)  4561.5(8)  «(Á)  8.3346(8)  13.1444(4)  19.614(2)  b(A)  8.4671(8)  9.1621(3)  17.114(2)  c(Á)  14.507(1)  16.7580(5)  15.519(2)  an  84.830(5)  90  90  85.483(5)  106.020(2)  118.880(4)  Crystal Data  X  0.20  X  0.20  rn z  68.227(4)  90  90  2  4  8  Density (calculated) (Mg/m^)  1.403  1.368  1.286  Absorption coefficient (mm'')  6.98  6.81  5.86  •^000  420  840  1872  Measured Reflections: Total  6332  17031  7526  Measured Reflections: Unique  3600  4608  3966  Final R índices''  R l = 0.0204, wR2  R l = 0.0244, wR2  R l = 0.0253, wR2  = 0.0535  = 0.0607  = 0.0591  1.068  1.031  1.021  0.729 and -0.365  0.308 and -0.250  Data Collection and Refínement  Goodness-of-fit o n i ^ *  Largest diff peak and hole (e" 0.354 and-0.456  " R l on F = S I (|Fo| - |Fc|) | / S |Fo|, ( / > 2a(7)); wR2 = [ (S ( F o ' - F e ' ) ' ) / S w(Fo' data); w = [ a^Fo" ]•'; * G O F = [ S (w ( |Fo| - |Fc| f ) / degrees offreedom  .  (all  Table 3.3 (continued) X-ray Crystallographic Data for Complexes 10,11 and 16. 10  11  16  Empirical formula  C23H37M0NO  C28H42M0N2O  C21H33M0NO  Crystal Habit, color  Needle, yellow  Block, orange  Irregular, yellow  Crystal size (mm)  0.5  0.40  0.50  Crystal system  Orthorhombic  Monoclinic  Monoclinic  Space group  Pbca  P2i/c  P2i/c  Volume (A^)  4337.3(10)  2648.2(6)  1997.1(4)  a (A)  8.8589(10)  14.8524(19)  8.934(1)  biA)  17.727(2)  9.2086(11)  13.776(2)  c(Á)  27.619(4)  20.619(3)  16.5414(8)  an fin rn z  90  90  90  90  110.106(7)  101.191(7)  90  90  90  8  4  4  Density (calculated) (Mg/m^)  1.346  1.301  1.368  Absorption coefficient (mm'')  6.16  5.17  6.64  Fooo  1856  1096  864  Measured Reflections: Total  7680  9679  7285  Measured Reflections: Unique  4201  5599  4157  Final R índices''  R l =0.0626, wR2 R l = 0.0338, wR2  R l = 0.0291, wR2  = 0.1274  = 0.0664  = 0.0713  1.120  0.922  1.069  0.421 and-0.337  0.440 and -0.445  Crystal Data  X  0.03  X  0.03  X  0.25 X 0.10  X  0.50  X  0.25  Data CoUection and Refinement  Goodness-of-fit on  *  Largest diff. peak and hole (e" 0.931 and-0.961  " R l on F = 2 I (|Fo| - \F,\) \ 12 \F^\, (/> 2a(/)); wR2 = [ (S ( Fo' - F / ) ' ) / 1 w(Fo' ff^ data); w = [ g^Fo^ ]"'; * GOF = [ S (w (|Foi - |Fc| f) / degrees offreedom f'^.  (all  3.5 References (1)  a) Wada, K.; Pamplin, C. B.; Legzdins, P.; Patrick, B. O.; Tsyba, I.; Bau, R. J. Am. Chem. Soc. 2003, J25, 7035-7048. b) Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, 223-233.  (2)  a) Graham, P. M . ; Buschhaus, M . S. A . ; Legzdins, P. J. Am. Chem. Soc. 2006,128, 90389039. b) Graham, P. M . ; Buschhaus, M . S. A . ; Pamplin, C. B.; Legzdins, P. Organometallics, 2008, 27, 2840-2851.  (3)  a) Christensen, Nancy J. Ph.D thesis, University of British Columbia, 1989. b) Hunter, A . D.; Legzdins, P.; Nurse, C. R.; Einstein, F. W. B.; Willis, A . C. J. Am. Chem. Soc. 1985, J07, 1791-1792. c) Christensen, N . J.; Hunter, A . D.; Legzdins, P. Organometallics, 1989, 8, 930-940.  (4)  Legzdins, P.; Rettig, S. J.; Veltheer, J. E.; Batchelor, R. J.; Einstein, F. W. B. Organometallics 1993,12, 3575-3585.  (5)  SIR92: Altomare, A . ; Cascarano, M . ; Giacovazzo, C ; Guagliardi, A . J. Appl Cryst. 1993, 26, 343.  (6)  Cromer, D. T.; Waber, J. T. International Tables for X-ray Crystallography; Kynoch Press: Birmingham, 1974; Vol. IV.  (7)  Ibers, J. A . ; Hamilton, W. C. Acta Crystallogr. 1964, 77, 781 -782.  (8)  Creagh, D. C ; McAuley, W. J. International Tables of X-ray Crystallography; Kluwer Academic Publishers: Boston, 1992; Vol. C.  (9)  Creagh, D. C ; Hubbell, J. H . International Tables for X-ray Crystallography; Kluwer Academic Publishers: Boston, 1992; Vol. C.  (10)  CrystalClear: Versionl.3.5b20; Molecular Structure Corporation, 2002.  (11)  SHELXL97: Sheldrick, G. M . University of Gottingen, Germany, 1997.  Chapter 4. Reactions of Cp'W(NO)(CH2CMe3)2 Complexes with Cyclic Olefins and Other Substrates^  A versión of this chapter has been submitted for publication. Buschhaus, M . S. A.; Pamplin, C. B.; Blackmore, I. J.; Legzdins, P. Transformations of Cyclic Olefins Mediated by Tungsten Nitrosyl Complexes. Reproduced in part with permission from Organometallics, submitted for publication. Unpublished work copyright 2008 American Chemical Society.  4.1 Introduction Both Cp*W(NO)(CH2CMe3)2 (1) and CpW(NO)(CH2CMe3)2 (2) are known to lose neopentane to genérate reactive alkylidene intermediates vmder theromlysis conditions. The Cp*alkylidene intermediate has been extensively studied in terms of its C - H bond activating abilities,' while the Cp-alkylidene intermediate has received less attention. Now, with oligomerization instead of C - H bond activation having been established as the preferred mode of reactivity with cyclic olefíns, the reactivity of 1 and 2 must be reevaluated. The investigations presented in this chapter seek to determine the extent of the cyclicolefin oligomerization ability of 1 and 2, and the organometallic products derived from 1 and 2 are examined in hopes of gaining insight into the oligomerization reaction mechanism that produces the organic products described in Chapter 2. The main part of this chapter examines the products formed when 1 is thermolyzed in cyclic olefíns with variable ring sizes and in a variety of cyclic substrates which either contain a methyl group or incorpórate a heteroatom in addition to the double bond. The last part of the chapter examines the reactivity of 2 (in contrast to 1) with regards to its tendency to undergo decomposition instead of C - H bond activation and its reactivity with the same cyclic olefín substrates. Unfortunately, for both CpW(NO)(CH2CMe3)2 and Cp*W(NO)(CH2CMe3)2, many reactions yield complex and often intractable mixtures of products. Nonetheless, some insights still emerge, and the characterizations of the isolable organometallic products contribute to subsequent mechanistic considerations.  4.2 Results and Discussion  4.2.1 Reactivity of 1 with cyclohexene: Evidence for the metallacycle 24 The organic oligomers formed during the thermolysis of 1 in cyclohexene can be isolated and characterized, at least as a mixture, as seen in Chapter 2. Conversely, the organometallic products have proved mostly intractable. In ' H N M R spectra of the crude product mixtures the peaks due to the oligomers obscure almost all of the peaks due to possible tungsten species. After the cyclohexene oligomers are removed by chromatography on an alumina column with pentane, repeated chromatographic attempts employing various ratios of pentane and Et20 as the eluting solvents afford only mixtures of organometallic species, without any effective separation. The fírst conclusión on viewing the ' H N M R spectra of these mixtures is that there are many organometallic products present. There are numerous peaks in the methyl región of the spectrum,  where signáis due to Cp* and neopentyl ligands are expected. No individual product can be isolated or defmitively characterized. At best some peaks can be grouped together as arising from the same compound and compared to similar systems. One tungsten compound in the mixture gives rise to singlet methyl resonances at 1.74 ppm (Cp*) and 1.22 ppm (CMes), to complex multiplets at -1.07 ppm and 0.09 ppm, and to a doublet at 3.64 ppm. The splitting pattems of the multiplets are similar to those of the molybdenum trans-metallacycles (9,14,17) discussed in Chapter 3 (see Table 3.1). Specifically, the molybdenum trans-metallacycle of cyclohexene (17) gives rise to ' H N M R peaks at -0.65 ppm (m), 0.1 ppm (ddd), and 3.91 ppm (d). The strong correlation in the relative shifts and in the splitting pattems of the characteristic multiplets between the molybdenum and tungsten systems suggests that a tungsten trans-metallacycle (24) is among the products formed in the reaction of 1 and cyclohexene (Scheme 4.1). This confírms that the tungsten alkylidene intermediate can indeed react with the double bond of cyclohexene, presumably foUowed by further reactivity and eventual decomposition.  Scheme 4.1  decomposition products  + organic oligomers 1  24  A l l in all, analysis of the organometallic products from the reaction of 1 with cyclohexene is only partially helpfiíl in terms of gaining insight into the oligomerization reactivity. The problem must be approached from a broader viewpoint. The first step involves the examination of other substrates, both to probé the extent and limits of the oligomerization reactivity of 1 and to seek clues about the oligomerization mechanism.  4.2.2 Reactivity of 1 with cyclopentene: The cis-metallacycle 25 Thermolysis of 1 in cyclopentene at 70 °C yields isolable organic and organometallic products (Scheme 4.2). The major organometallic product precipitates as an orange solid from a cooled EtiO solution of the reaction residue. The products remaining in solution are then chromatographed on alumina, and the organic products are recovered as clear, colorless oil from the pentane fraction. Organometallic decomposition products are recovered in the bronze-brown EtaO fraction.  Scheme 4.2  oligomeric products  1  25  The orange solid has been identified as the metallacyclobutane 25,^^ a direct analogue of the molybdenum cis-metallacycle 7. The cyclopentene has undergone a 2 + 2 addition across the metal-carbon double bond of the alkylidene to form the metallacycle. Under the conditions described above, 25 forms in about 76% yield. The ' H N M R spectrum contains diagnostic multiplet signáis arising from the three protons on the carbons of the metallacycle ring. The ahydrogen on the cyclopentane ring gives rise to a downfield doublet-of-triplets centered at 7.65 ppm, and the P-hydrogen signal appears as a complex multiplet at -0.48 ppm. The signal of the a-hydrogen nearest the *Bu group appears as a doublet at 3.18 ppm. The formation of 25 demonstrates conclusively that the alkylidene M=C bond can indeed react with the alkene functionality of the cyclic olefin to form the predicted metallacyclobutane. However, complex 25 does not shed much insight on the oligomerization reactivity of interest. As expected when using a five-membered cyclic substrate, the formation of 25 suggests the initial step of coupling a neopentyl imit to a cyclic olefin but reveáis nothing with regard to the identity of any catalytically active tungsten species.  The organic products formed from cyclopentene conform to the general characteristics previously discussed for the cyclohexene oligomers. Low-resolution mass spectral analyses consistently show oligomer lengths in the range of dimers to undecamers or dodecamers. Unsaturation levéis increase with oligomer length, with typically two unsaturations in trimers to septamers, three in octamers to undecamers, and as many as four in dodecamers. The peaks between 5.3 and 5.8 ppm in the ' H N M R spectrum confírm that olefmic hydrogens remain in the oligomers. A G C / M S analysis yields a complex GC trace, in which peaks for dimers through septamers can be identified. However, due to the large number of overlapping and thus unassignable peaks, no integration or other analysis can be done. The yield of cyclopentene oligomer per mole of precatalyst 1 is lower than the yields obtained for cyclohexene. Calculated according to the entire amount of precatalyst 1 used, the conversión is only 5.6 moles of cyclopentene per mole of 1. However, taking into account the amount of 25 recovered, which has effectively delayed three-quarters of the tungsten from reaching the catalytic cycle, the tumover number is 2 4 moles of cyclopentene per mole of active tungsten species. Still, this conversión is not as high as those obtained with cyclohexene (up to 83 moles of cyclohexene converted per mole of tungsten at 70 °C).  4.2.3 Reactivity of 1 with cyclooctene Thermolysis of 1 in cyclooctene at 70 °C yields a mixture of products, which ' H N M R analysis reveáis to include at least three major organometallic compounds and some organic products. Column chromatography on alumina isolates the organic products in the pentane fraction. Two of the organometallic products elute in the EtaO fraction, but the third organometallic complex does not elute from the column. The structure of the product that cannot be isolated is proposed to be a metallacycle (26) by ' H N M R analysis. The two isolable organometallics are separated by selective precipitation, and solid-state X-ray analyses reveal their structures to be 27 and 28. Scheme 4.3 summarizes the products obtained in the reaction.  4.2.3.1 Thefirstcyclooctene product: Evidence for the trans-metallacycle 26 The identity of compound 26 is deduced by comparison of its diagnostic ' H N M R peaks, as observed in the spectrum of the crude reaction mixture, with those of the analogous molybdenum compound. Specifically, the three hydrogens of the metallacycle in 26 give rise to distinctive peaks at -0.49 ppm, 3.98 ppm and 4.03 ppm. The splitting patterns are dq, dt, and d, respectively. The fuUy characterized molybdenum analogue (9) gives rise to peaks with similar shift and splitting pattems (see Table 3.1). Thus, like compound 24 above, compound 26 suggests a coupling of the cyclic olefin substrate with the alkylidene intermediate of 1 to form a metallacycle and provides a link between the observed tungsten and molybdenum reactivities.  4.2.3.2 The second cyclooctene product: The 1,4-diene 27 Compound 27 readily precipitates as a red solid from a cold Et20 solution. Mass spectral analysis gives a parent mass of 567 amu, indicating the loss of the neopentyl ligands and the inclusión of two cyclooctene units at the metal center. The solid-state molecular stmcture confirms these changes and reveáis the new ligand to be two coupled cyclooctene units connected to the metal center as a 1,4-diene (Figure 4.1). The solid-state molecular stmcture  contains disorder at C14 and C15, which has been modeled as two parts that contribute 70% and 30%, respectively. A s a result of this disorder and the effect of the small size of the crystal on the data set, the final R valué is 6.8%), but the important structural details are unambiguous.  Figure 4.1 Solid-state molecular structure of 27 (part A) with 50 %> probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W ( l ) - C ( l ) = 2.283(10), W ( l ) C(8) = 2.458(13), W(l)-C(10) = 2.264(13), W(l)-C(l 1) = 2.217(13), W(l)-N(l) = 1.772(9), N(l)-0(1)= 1.238(12), C ( l ) - C ( 8 ) = 1.390(17), C(8)-C(9) = 1.529(19), C(9)-C(10) = 1.505(16), C ( 1 0 ) - C ( l l ) = 1.431(17), C(l)-C(8)-C(9)= 121.8(12), C(8)-C(9)-C(10) = 101.9(10), C(9)C(10)-C(l 1) = 125.3(11), W(l)-N(l)-0(1) = 174.8(9).  ' H and ' ^ C N M R analyses of 27 confirm the presence of the Cp* ligand with a protón singlet at 1.65 ppm and the absence of neopentyl groups by the lack of other strong singléis. Not surprisingly, the coupled cyclooctene ligand gives rise to many overiapping signáis. The diagnostic peaks of the hydrogen and carbón atoms nearest the metal center can be readily  assigned, but those of the CH2 groups are difficult to differentiate. The carbón signal of C l appears at 72.1 ppm, and the related protón signal appears as a doublet of doublets at 2.14 ppm. The quatemary carbón C 8 gives a peak at 69.4 ppm. The carbons of the other coordinated olefin give rise to peaks at 42.3 ppm (CIO) and 47.5 ppm ( C l 1). The olefinic protón signáis appear as multiplets at 4.48 ppm (HIO) and 0.90 ppm (Hl 1). Finally, C 9 gives rise to a peak at 9.5 ppm and H9 to a triplet at -1.26 ppm. Monitoring the thermolysis of 27 in CeDó at 70 °C over four days by ' H N M R spectroscopy shows the decrease and almost complete loss of signáis for 27. Peaks due to 28 appear and then eventually decrease with increasing time, showing that 27 is able to convert to 28 under thermolysis conditions, but then decomposes. Concomitant with the loss of 27 and 28 is the appearance of at least two new multiplet peaks in the olefinic región (5.35 ppm and 5.48 ppm). These are attributed to the reléase of the coupled cyclooctene as an organic molecule, presumably l-cyclooct-3-enylcyclooctene, although unfortunately the splitting pattems of the peaks are not sufficiently resolved to analyze ñiUy. Peaks attributed to decomposition products grow in over the four day reaction time. Specifically, the largest ones in the Cp* región are at 1.47 ppm, 1.71 ppm and 1.72 ppm, in addition to many smaller singlets between 1.3 ppm and 2.2 ppm. The identity of these products has not been ascertained. Thermolysis of 27 in cyclohexene at 70 °C for 24 h results in formation of oligomers of cyclohexene, as identified by their characteristic peaks in the ' H N M R spectmm of the cmde reaction mixture (Figure 4.2). Characteristic peaks for 27 and 28 appear in an approximate 1:5 ratio, confirming that 27 converts to 28 under thermolysis conditions in addition to forming the catalytically-active species of the oligomerization cycle. Finally, strong singlets at 1.52 ppm and 1.48 ppm suggest the formation of tungsten decomposition products over time.  I  27  "T^^^-^T-T^ '  7.0  e.5  I  6.0  5.5  III.I IIIIII  'I  5.0  4.S  4.0  i I I . i MT^. I I I I, I , I, I. I I I I > M ' • " I " " I i • " I " " I " i • I ' " M '  3.5  3.0  2.5  2.0  l.S  1.0  0.5  IiMI, II  O  .0.5  , r, I . I , I I n  -1.0  -1.5  Figure 4.2 ' H N M R spectrum (300 MHz, CgOe, rt) of the product mixture from thermolysis of 27 in neat cyclohexene at 70 °C for 24 h, demonstrating the formation of cyclohexene oligomers, the formation of 28 and the remaining presence of unreacted 27. The arrows indícate diagnostic peaks due to 27 and 28, while the cyclohexene oligomers give rise to broad peaks between 1 - 2 ppm and 5.3 - 5.8 ppm.  4.2.3.3 The third cyclooctene product: The allyl hydride 28 The other product isolated from thermolysis of 1 in cyclooctene, 28, recrystallizes as a yellow prism. Its parent mass is also 567 amu, and the solid-state molecular structure (Figure 4.3) reveáis that the coupled cyclooctene ligand is coordinated to the metal center through an allyl linkage. The metal center must also support a hydride ligand, as evidenced by the parent mass and the W-H signal observed by ' H N M R spectroscopy (vide infra). Unfortunately, the hydride could not be specifically located in the residual electrón density map, despite the vacant coordinatíon site on tungsten in the solid-state molecular structure. The coupled cyclooctene ligand contains an uncoordinated double bond between CIO and C l 1. Notwithstanding disorder in the Cp* ligand, the final R valué for the structure is 3.5%. The structures of 27 and 28, as well as their ability to interconvert, suggest that the complexes are related by way of a reversible (3hydrogen C - H activation.  W1  Figure 4.3 Solid-state molecular structure of 28 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.279(3), W(l)-C(8) = 2.335(3), W(l)-C(9) = 2.472(3), W(l)-N(l) = 1.767(3), N(l)-0(1) = 1.229(4), C(l)-C(8) = 1.445(4), C(8)-C(9)= 1.422(4), C(9)-C(10) = 1.498(4), C(10)-C(l 1) = 1.353(5), C(l)-C(8)-C(9) = 118.0(3), C(8)-C(9)-C(10)= 118.0(3), C(9)-C(10)-C(l 1) = 130.9(3), W(l)-N(l)-0(1) = 169.7(3). Hydride not located.  The N M R data for 28 indícate that the solid-state molecular structure is retained in solution. hl the ' H N M R spectrum the Cp* ligand gives rise to a strong singlet at 1.76 ppm. The olefínic protón signáis of HIO and H l 1 appear as a doublet at 7.00 ppm and a doublet of doublets of doublets at 5.17 ppm, respectively. The signal for the allyl protón ( H l ) is buried under múltiple methylene peaks around 1.55 ppm. In the '^C N M R spectrum, the allylic carbons are indicated by peaks at 64.0 ppm (Cl), 120.8 ppm (C8) and 91.2 ppm (C9). The carbón signáis  of the free olefin appear at 138.8 ppm (CIO) and 119.1 ppm ( C l 1). Signáis due to the methylene groups of the coupled ligand are difficult to assign specifically. In the '^C N M R spectrum, eleven peaks appear between 20.9 ppm and 35.0 ppm. In the ' H N M R spectrum, half of the methylene protons are represented by overlapping multiplets in the regions of 1.32 ppm and 1.55 ppm. Signáis for the methylene groups next to the allyl and olefin are spread over a greater range, with multiplets that intégrate for one, two, or three protons appearing at 2.92 ppm, 2.44 ppm, 2.22 ppm, 2.03 ppm, 1.93 ppm, and 0.54 ppm. Finally, the hydride signal appears as a singlet at 0.12 ppm flanked by satellites with a VWH coupling of 124.6 Hz.  4.2.3.4 The organic products of cyclooctene The oligomeric products originally formed during the thermolysis of 1 in cyclooctene exhibit characteristics similar to those observed for the cyclohexene oligomers. ' H N M R analysis shows complex peaks in the aliphatic región between 0.8 and 3.3 ppm and in the olefinic región between 5.2 and 5.8 ppm. At least two of the multiplets in the downfield región are obviously doublet-of-triplet signáis, suggesting olefinic protons next to a CH2 group, but beyond this observation specific isomers cannot be identified in the mixture. Integration indicates that the aliphatic to olefinic protón ratio is 9:1. Since a coupled dimer with one unsaturation would give a ratio of 13:1 and a dimer with two unsaturations would give 5.5:1, the experimental data suggest an average of one to two unsaturations per two cyclooctene units. Low-resolution mass spectrometric analyses (with varying temperature conditions) show mass peaks for dimers and trimers with two unsaturations, and for higher oligomers from tetramers to septamers with one unsaturation. A GC/MS analysis produces a G C trace containing peaks due to cyclooctene coupled to neopentyl with one or two unsaturations, cyclooctene dimers capped with neopentyl, and cyclooctene dimers and trimers with two unsaturations. Múltiple peaks appear for dimers and trimers, suggesting several isomers are present. Overall, the organic products consist of shorter oligomers than those formed from cyclohexene, but the ring-retaining structure is analogous. In a reaction with an initial concentration of 0.01 M thermolyzed at 70 °C, the tumover is calculated at 1.45 moles of cyclooctene converted to oligomer per mole of precatalyst 1. (The recovered amounts of 27 and 28 were not subtracted from the moles of tungsten catalyst since they may form before or after oligomerization activity.) The conversión ratio is small compared to valúes obtained for cyclohexene and cyclopentene. It seems that, as for the molybdenum system, the large ring size of cyclooctene is slowing the reactivity.  In terms of mechanistic insight, the proposed structure of compound 26 demonstrates the interaction of the cyclooctene substrate with the alkylidene intermediate of 1. Compounds 27 and 28 represent some later stage in the reaction mechanism, with the complete loss of the neopentyl ligands and the coupling of two cyclooctene molecules in the coordination sphere of the metal. The GC/MS analysis of the organic products demonstrates both the coupling of cyclooctene to a neopentyl unit and the coupling of cyclooctene with itself. These key insights will be fuUy considered in the mechanistic discussions of Chapter 5.  4.2.4 Reactivity of 1 with 4-methylcycIohexene The successful oligomerization of cyclic olefms prompts attempts to extend this reactivity to other substrates. The presence of a methyl group in 4-methylcyclohexene allows for examination of a C - H bond activation reaction of a primary carbón that could compete with oligomerization. The methyl group also adds some bulk to the substrate but not in a position near the double bond. Precatalyst 1 reacts with 4-methylcyclohexene under thermolysis conditions to produce organic oligomers, but in far lower yields than the related cyclohexene system. Thus, after thermolysis at 70 °C, an initial 100 mg of 1 affords only 8 mg of 4-methylcyclohexene oligomer, and the tumover is 0.4 moles of 4-methylcyclohexene converted per mole of 1. (Recall that similar cyclohexene reactions produced gram amounts of oligomer.) The isolated 4-methylcyclohexene oligomers contain unsaturated positions (double bonds), as evidenced by the broad peaks in the olefinic región between 5.2 and 5.8 ppm in the ' H N M R spectrum. As with the cyclohexene oligomers, peaks due to the 4-methylcyclohexene oligomer families appear as clusters in the G C trace. Specifically, four peaks due to dimers capped with a neopentyl unit appear at 7.3-7.6 min, a broad lump of many peaks due to trimers appears at 9.9-11.3 min, six peaks due to trimers capped with a neopentyl unit appear at 12.413.3 min, and many peaks due to tetramers appear at 14.1-14.9 min. G C evidence for 4methylcyclohexene dimers is negligible. The baseline of the G C trace rises between 17 and 18 min, but this cannot be conclusively assigned to the presence of pentamers. Integration of the assignable G C peaks gives a relative mixture composition of 9% neopentyl-capped dimers, 72% trimers, 3% neopentyl-capped trimers, and 16% tetramers. The M S fragmentation pattems of the oligomers are consistent with retention of the ring, showing the sequential loss of 7-carbon units. The parent masses in each family show that the  unsaturation levéis increase with oligomer length. Thus, neopentyl-capped dimers have one unsaturation, trimers and neopentyl-capped trimers have two unsaturations each, and tetramers vary between two and three unsaturations. The presence of the methyl group on the cyclic olefin affects the oligomerization reaction. First, as noted above, the overall substrate conversión and oligomer yield drop drastically. Second, as shown in the G C data, the distribution of the oligomers shifts to a higher percentage of trimers and lower percentages of longer oligomers. The percentages of oligomers capped with a neopentyl group are also high, likely because the low conversión leads to a seeming enrichment for the initiation-derived products. Third, the number of unique oligomer compounds increases as evidenced by an increased number of individual peaks in the GC peak families. Thus, four peaks appear for dimers capped with a neopentyl unit, whereas in the cyclohexene-oligomer G C trace there is only one peak. Mechanistically, this increase in the number of isomers can be accounted for i f the neopentyl unit is coupled in various positions relative to the methyl group of the substrate. The organometallic products produced in the reaction of 1 with 4-methylcyclohexene form a complex mixture. ' H N M R analyses of both the crude reaction mixture and the organometallic products isolated by chromatography (Et20 fraction) show large numbers of peaks in the methyl región of the spectrum, where the diagnostic protón signáis of the Cp* ligand and of the neopentyl ligand are expected to appear. None of the organometallic products can be individually isolated.  4.2.5 Comparison of turnover numbers for the cyciic-olefín substrates with 1 As discussed in the proceeding sections, precatalyst 1 oligomerizes the cyclic olefins to varying extents. Table 4.1 summarizes the tumover numbers and the oligomer distributions obtained for cyclohexene, cyclopentene, cyclooctene and 4-methylcyclohexene. It is clear that cyclohexene is the best substrate for oligomerization.  Table 4.1 Comparison of the cycUc-olefin substrates oUgomerized by precatalyst 1 based on tumover number (TON) and oligomer distribution. Substrate  Cyclohexene  Oligomer distril )ution (]Dercenta ge by GC)"  TON'  83,' 149"  olefin + npt  dimer  -  7  Cyclopentene  24'  Cyclooctene  1.45'  6  4-methylcyclohexene  0.4  -  dimer + npt  1 <1  i  trimer  trimer + npt  tetramer  45  _  36  pentamer & higher  i  11  not assigned i  90  i  2  1  9  72  1  i  3  !  16 1  1  ^TON = moles substrate converted per mole 1;" olefin = cyclic-olefm substrate, npt = neopentyl end group; "best result at 70 °C; ''at 100 °C; "calculated based on active catalyst formed; recovered amounts of 26-28 not subtracted.  4.2.6 Reactivity of 1 with oxygen-containing cyclic olefíns In order to further explore the functional group tolerance of the oligomerization reaction, three substrates containing oxygen have been thermolyzed in the presence of 1 (Scheme 4.4).  Scheme 4.4  2,5-dihydrofuran  2,3-dihydrofuran  3,4-dihydro-2H-pyran  4.2.6.1 Thermolysis of 1 in 2,5-dihydofuran: The cis-metallacycle 29 Thermolysis of 1 in 2,5-dihydrofuran at 70 °C yields the metallacycle 29 as the main product (Scheme 4.5). Compound 29 can be recrystallized directly from an Et20 solution of the cmde reaction products, but it is destroyed on an alumina column. The dark-colored minor products that remain in solution have not been identified. No evidence for organic products has been detected.  The soHd-state molecular structure of 29 shows the cis-metallacycle configuration (Figure 4.4, next page). The tungsten center is almost co-planar with C2, C3 and C5, with a torsión angle of 3.23°. The protons on C2 and C3 are positioned cis to each other and downward relative to the Cp* ligand. The protón on C5 (not shown) points upward in the direction of the Cp* ring. The final R valué for the structure is 1.6%. N M R analyses are consistent with the 2 + 2 metallacyclic structure, with the signáis for the carbón and hydrogen atoms of the tungstenacycle being particularly diagnostic. In the '^C N M R spectrum C2 and C5 give rise to peaks at 117.3 ppm and 128.9 ppm, respectively, while C3 gives rise to a peak at 15.1 ppm. In the ' H N M R spectrum, H3 is attributed to a resonance at -0.46 ppm, H2 to a peak at 6.67 ppm, and H5 to a peak at 3.75 ppm. These diagnostic N M R valúes for 29 are comparable to those of complex 25 (vide supra) and those of the molybdenum complex 7. In all three cases, the cyclic-olefin derived a-hydrogen of the cis-metallacycle gives rise to a downfield signal while the P-hydrogen gives rise to an upfield signal. The solid-state molecular structure of 29 also compares favorably with that of 7. Both structures show the essentially planar metallacycle, and in both the two cis protons are downward relative to the Cp* ligand and the coupled ring.  W1  Figure 4.4 Solid-state molecular structure of 29 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W ( l ) - C ( 2 ) = 2.079(2), W ( l ) - C ( 5 ) = 2.104(2), C(2)-C(3) = 1.659(3), C(3)-C(5) = 1.586(3), W(l)-C(2)-C(3) = 77.68(11), W ( l ) - C ( 5 ) C(3) = 78.44(12), C(5)-C(3)-C(2) = 119.55(17), C(2)-W(l)-C(5) = 84.22(9), W ( l ) - C ( 2 ) - C ( l ) = 127.94(16), C(5)-C(3)-C(4) = 113.67(19), C(2)-C(3)-C(5)-W(l) = 3.23(17).  4.2.6.2 Thermolysis of 1 in 2,3-dihydofuran In contrast, thermolysis of 1 in 2,3-dihydroñiran at 70 ° C yields an intractable mixture of products. After removal of the volátiles in vacuo, the reaction residue is dark and oily. Extraction of this residue with pentane and with Et20 does not allow isolation of any individual products. The ' H N M R spectra have many peaks in the methyl región, indicating a multitude of compounds. There are also overlapping multiplets almost continuously from 2 to 6 ppm. Among these peaks, there is no indication of a major organometallic product. In addition, comparison of ' H N M R spectra does not defmitively identify any signáis analogous to those seen for oligomers.  4.2.6.3 Thermolysis of 1 in 3,4-dihydro-2H-pyran: The alkoxy allyl 30 Thermolysis of 1 in 3,4-dihydro-2H-pyran at 7 0 °C yields a complex mixture of products. The ' H N M R spectrum of this mixture shows múltiple broad peaks between 0.7 and 2.3 ppm, out of which several large singlets rise. In addition, many small peaks appear between 3 and 6 ppm. Chromatography on alumina followed by múltiple extractions affords a yellow solution, from which a single product can be isolated as a palé powder. Analysis of this product by ' H , ' ^ C ' H ' H COSY, ' H - ' ^ C H M B C , ' H - ' ^ C H M Q C , and '^C A P T N M R experiments identifíes it as the ring-opened alkoxy allyl 30 illustrated in Scheme 4.6.  Scheme 4.6  other products  1  30  Specifícally, in the ' H N M R spectrum of 30 the Cp* ligand and the CMca group give rise to strong singlets at 1.67 ppm and 1.18 ppm, respectively. The central protón on the allyl gives rise to a downfíeld peak at 4.35 ppm, and the signáis from the two outer allyl protons overlap at 2.34 ppm. The protons of the methylene group a to the oxygen give rise to peaks at 3.52 ppm and 3.95 ppm, and those p to the oxygen to multiplets at 1.41 ppm and 1.97 ppm. The signal at 1.97 ppm overlaps with the signáis of the methylene group positioned next to the tert-butyl group. In the '•'c N M R spectrum, signáis at 59.2 ppm, 81.5 ppm and 42.5 ppm are attributable to the allyl carbons. The methylene a to the oxygen gives a peak at 62.9 ppm, and the other methylene carbons give peaks at 29.3 ppm and 32.1 ppm. Analysis of the ' H N M R spectrum of the crude reaction mixture shows that 30 is one of the major products formed. The other products have not been identifíed, ñor has a crystal of 30 suitable for X-ray analysis been obtained.  4.2.6.4 Comparisons to the reactivity of 5 with oxygen-containing cyclic olefíns The structural formulation of 30 shows similarities to the reaction product obtained from thermolysis of 5 in the same cyclic olefin substrate. Specifically, 5 has been thermolyzed in 3,4dihydro-2H-pyran and in 2,5-dihydrofuran (by Chris Semiao and Dr. lan Blackmore) to yield the alkoxy products 31 and 32, respectively, as illustrated in Scheme 4.7. In all three cases where the ring opens, complexes 1 and 5 presumably first lose neopentane or tetramethylsilane to genérate their respective reactive intermediates (an alkylidene or an ri'-alkyne), which then react with the cyclic-olefin substrates. The net result is cleavage of the C - 0 bond of the cyclic substrate, accompanied by coupling and rearrangement with the reactive intermediate fragment. The solid-state molecular structures of 31 and 32 have been determined by single crystal X-ray analyses and are shown on the following pages in Figure 4.5 and Figure 4.6, respectively. The final R valúes of the structures are 1.6% and 2.5%. In particular, compound 31 shows an alkoxy allyl binding motif comparable to that of 30. Key ' H and '^C N M R data for 30 and 31 include the downfield resonances due to the methylene protons a to the alkoxy oxygen and the expected coupling patterns of the ri^-allyl protón signáis.  Scheme 4.7  \  32  Figure 4.5 Solid-state molecular structure of 31 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-0(2) = 2.0163(17), W(l)-C(3) = 2.423(3), W(l)-C(4) = 2.303(3), W(l)-C(5) = 2.284(3), 0(2)-C(l) = 1.398(3), C(l)-C(2) = 1.530(4), C(2)-C(3) = 1.479(4), C(3)-C(4) = 1.368(4), C(4)-C(5) = 1.439(4), C(5)-C(6) = 1.452(4), C(6)-C(7) = 1.328(4), W(l)-0(2)-C(l) =117.69(16), W(l)-N(l)-0(1) = 167.9(2).  In comparing the four products isolated from the reaction of 1 and 5 with oxygencontaining heterocycles, both the ring size and the position of the double bond in the substrate seem to díctate whether and how the ring will open to give the observed products. In one case the five-membered 2,5-dihydrofiiran ring is opened, while in the other it is not. The sixmembered 3,4-dihydro-2H-pyran is opened in both cases to give similar alkoxy allyl complexes. However, in terms of the oligomerization reactivity, these substrates are disappointing. No direct evidence for oligomers is obtained from the reaction mixtures. Thus, the oxygen functional group within the substrate ring leads to a divergent reaction mechanism that favors a C - 0 bond opening reaction over the desired ring-retaining oligomerization.  Figure 4.6 Solid-state molecular structure of 32 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-0(2) = 2.097(2), W(l)-C(6) = 2.209(3), 0(2)-C(l) = 1.421(4), C(l)-C(2) = 1.488(5), C(2)-C(3) = 1.497(5), C(3)-C(4) = 1.296(6), C(2)-C(5) = 1.485(5), C(5)-C(6) = 1.335(5), W(l)-0(2)-C(l) = 117.2(2), W(l)-N(l)0(1)= 170.3(3).  4.2.7 Reactivity of 1 with nitrogen-containing cyclic olefíns Substrates that include an amine group within the ring have also been thermolyzed with precatalyst 1. These substrates, 1,2,3,6-tetrahydro-pyrídine and 3-pyrroline, are illustrated in Scheme 4.8.  \ /  N í H  1,2,3,6-tetrahydro-pyridine  NH  3-pyrroline  4.2.7,1 Thermolysis of 1 in 1,2,3,6-tetrahydro-pyridine: The amido complex 33 Thermolysis of 1 in 1,2,3,6-tetrahydro-pyridine at 70 °C produces one major organometallic product (Scheme 4.9). Isolation and recrystallization from EtaO yield bright orange crystals of 33 in high yield. No oligomeric products are detected by ' H N M R spectroscopy.  Scheme 4.9  1  33  A n X-ray crystallographic analysis of 33 shows that the tungsten center is in the same plañe as the nitrogen and a-carbons of the substrate ring, with the torsión angle of C(l)-N(2)C(5)-W(l) equal to approximately 8.6°, The double bond of the substrate ring is shown between C2 and C3 for the purposes of generating an ORTEP picture (Figure 4.7). However, the crystallographic data actually represent an averaged form where the olefin can be in one of two possible positions, C2=C3 or C3=C4, as evidenced by the similar C2-C3 and C3-C4 bond lengths in the structure. The bond lengths of C1-C2 and C4-C5 are also similar, as expected for these single bonds. The final R valué is 2.6 %.  Figure 4.7 Solid-state molecular structure of 33 with 50 % probabihty thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(6) = 2.183(4), W(l)-N(2) = 1.939(3), W(l)-N(l) = 1.765(3), N(l)-0(1)= 1.233(4), N(2)-C(l) = 1.469(5), C(l)-C(2) = 1.494(7), C(2)-C(3) = 1.404(7), C(3)-C(4) = 1.393(7), C(4)-C(5) = 1.484(6), C(5)-N(2) = 1.465(5), W(l)-C(6)-C(7) = 123.8(3), W(l)-N(2)-C(l) = 123.3(3), C(l)-N(2)-C(5) = 109.5(3), W(l)-N(l)-0(1)= 168.3(3).  The ' H N M R spectrum of recrystallized 33 in CeDg shows peaks arising from two isomers, present in a 3:1 ratio. The major isomer gives rise to a Cp* signal at 1.65 ppm, a neopentyl methyl resonance at 1.40 ppm, and a neopentyl methylene doublet signal at 0.97 ppm. The minor isomer signáis for the same ligands appear at 1.64 ppm, 1.41 ppm and 1.04 ppm, respectively. The resonances due to the amido ligand appear as broad multiplets with extensive fine structure. The olefinic protón peaks for both isomers overlap and appear at 5.23 ppm and  5.51 ppm. The resonances due to the aliphatic protons show isomer dependent differences in shift. Specifically, for the major isomer the peaks appear at 1.77 ppm (H4), 2.00 ppm (H4), 3.44 ppm (Hl), 3.90 ppm (H5), and 4.98 ppm (H5). In contrast, for the minor isomer the two NCH2C//2CH=CH  peaks are obscured in the región where the signáis attributed to  H4  of the  major isomer appear. The remaining peaks appear at 3.17 ppm (NC//2CH=CH), 4.30 ppm (NC7/2CH2),  and 5.37 ppm  (NC//2CH2).  The relative changes in shift between peaks for the two  isomers, particularly those associated with the protons on the carbons a to nitrogen, suggest a mixture of isomers in which both possible locations of the double bond within the ring relative to the rest of the complex are represented. Thus, in one isomer the protons on C l would be cis to the neopentyl ligand, while in the other isomer the position of C l within the ring would be cis to the nitrosyl ligand. This is consistent with the solid-state molecular data showing an averaged form of these two isomers. It is not possible to state definitively which isomer corresponds to which orientation, although for the purposes of discussion the major isomer is assumed to match the orientation shown in Figure 4.7. Chromatography of the pentane-soluble products through an alumina column does not isolate any oligomeric products. Compound 33 is thermally stable. After the complex is heated at 100 °C in CeDó for a week, ' H N M R analysis indicates that only slight decomposition of 33 has occurred. Thus, 1,2,3,6-tetrahydro-pyridine is not oligomerized by 1, ñor does a metallacycle form by addition of the olefin functionality to the alkylidene. Rather, the alkylidene intermediate seems to have initiated N - H bond activation to produce 33 in high yield. Complex 33 cannot serve as a precursor to the oligomerization reactivity of interest, as it is thermally stable.  4.2.7.2 Thermolysis of 1 in 3-pyrroline Thermolysis of 1 in 3-pyrroline at 70 °C yields an intractable mixture of products. Monitoring the reaction by ' H N M R spectroscopy suggests that impurities in the substrate complícate the reactivity. This complex reactivity is in sharp contrast to the formation of cismetallacycles from pentene (25) and 2,5-dihydrofuran (29), and to the N - H bond activation achieved with 1,2,3,6-tetrahydro-pyridine to yield 33. Whether this reactivity with 3-pyrroline is due solely to impurities in the substrate or also to the substrate itself cannot be judged. No evidence for oligomerization products appears in ' H N M R or mass spectral data.  4.2.8 A brief summary of the isolable Cp* tungsten organometallic products Compound 1 reacts with the double bond of cyclic olefín substrates. A fíve-membered ring forms the cis-metallacyclobutane 25, while there is evidence that the larger ring sizes form trans-metallacycles (24 and 26). With cyclooctene, two additional products are isolated which contain coupled cyclooctene ligands, the fírst a 1,4-diene complex (27) and the second an allyl hydride complex (28). With oxygen-containing substrates, fíve-membered 2,5-dihydrofuran forms the cis-metallacycle 29, while six-membered 3,4-dihydro-2H-pyran forms many products, one of which is identified as an alkoxy allyl complex (30). With nitrogen-containing substrates, a clean reaction occurs with 1,2,3,6-tetrahydro-pyridine to produce 33 by N - H bond activation.  4.2.9 Reactivity of 2: Trapping the alkylidene intermediate with trimethylphosphine Like compound 1, compound 2 forms an alkylidene intermediate under thermolysis conditions. The alkylidene intermediate thus formed can be trapped with trimethylphosphine as CpW(NO)(=CHMe3)(PMe3) (34), as fírst observed by Elizabeth Tran.^ Complex 34 is the major product isolated from thermolysis reactions of 2 in the presence of PMea at 60 °C when THF or cyclohexene are used as solvents. In contrast, the thermolysis of 2 in cyclohexane in the presence of PMe3 results in the formation of two products, 34 and CpW(NO)(C6Hio)(PMe3) (35), as shown in Scheme 4.10.  Scheme 4.10  2  34  35  4.2.9.1 The decomposition products of 2 in cyclohexane Having estabUshed the formation of the alkylidene intermediate under thermolysis conditions, the ability of 2 to actívate C - H bonds has been explored with various substrates. For the most part, 2 does not form products analogous to those formed from 1 under similar conditions.' Rather, 2 tends to form solvent-independent decomposition products. The main decomposition products of 2 form most cleanly when cyclohexane is used as the solvent. Thus, when 2 is thermolyzed at 6 0 °C in cyclohexane for 4 0 h, four major organometallic products form, as evidenced by four distinct peaks due to Cp in the ' H N M R spectrum of the crude product mixture. These four products, called 36, 37, 38, and 39 for the sake of discussion, are difficult to sepárate and characterization is incomplete. Low-resolution mass spectrometric analyses indícate that all four products are dimeric species. On the basis of available ' H N M R data, empirical formulae are suggested. Complex 36 is formulated as [CpW(NO)(CH2CMe3)2]2, suggesting that 2 has symmetrically dimerized without loss of neopentane or formation of the alkylidene intermediate. Compound 37 has lost two hydrogen atoms to form the symmetric dimer [CpW(NO)(CH2CMe3)(=CHCMe3)]2. Compound 38 is the simplest dimer, [CpW(NO)(CH2CMe3)]2, with the neopentyl groups in symmetric environments. Compound 39 forms in small amounts and its characteristic data indícate that it containing two equivalent Cp ligands and two inequivalent alkyl ligands. Under certain high concentration conditions, a fifth complex (40) forms that contains two inequivalent Cp ligands and two inequivalent alkyl ligands. Monitoring the further thermolysis of 36 and 37 in CeDg by ' H N M R spectroscopy reveáis the loss of 36 and the formation of 38. As the signáis due to 36 disappear and those of 38 grow in, neopentane is detected in the product mixture as evidenced by a singlet at 0.80 ppm in the ' H N M R spectrum, thus indicating that 36 converts to 38 with loss of neopentane. It is harder to tell if 37 converts in a similar manner. Compound 38 is the final thermodynamic product. The room temperature reaction of 2 in cyclohexane affords the decomposition products 36, 37 and 38 in equal amounts after 3 4 days, as well as unreacted 2. No other reaction products are observed at rt, indicating that the products obtained under thermolysis conditions are not merely the result of thermally unstable initial products. When thermolysis of 2 is performed in cyclohexane there is no evidence for alkylidene-based reactivity, and the decomposition products 36-40 indícate that most probably 2 reacts with itself  4.2.9.2 Reactivity of 2 with other substrates Compound 2 has been thermolyzed in a variety of substrates in order to explore its C - H bond-activating ability. Generally, the thermolyses have been carried out at 60 °C, and the number of products in the final reaction mixtures is estimated on the basis of the number of peaks attributable to Cp in the ' H N M R spectrum in the región between 5 and 6 ppm. Thermolysis of 2 in tetramethylsilane results in more than ten products, including 37 and 38. Thermolysis of 2 in mesitylene results in about eight products, including large amounts of 36 and 38 and small amounts of 37 and 39. There is no clear evidence for C - H bond activation reactivity. Thermolysis of 2 in dichloromethane results in more than ten peaks due to Cp in the ' H N M R spectrum, but none of these could be identified at all. Thermolysis of 2 in the solid state also produces decomposition products, the three major ones being identified as 36, 37 and 38. Therefore, whether in solution or as a solid, decomposition pathways predomínate, as summarized in Table 4.2.  Table 4.2 Summary of substrates reacted with 2 and the number of products formed. Temperature (°C)  Reaction Time  Cyclohexane  60  41 h  Cyclohexane  rt  34 days  Tetramethylsilane  40  16.5 h  Mesitylene  60  22 h  Dichloromethane  60.  69 h  > 10 - imidentified  Solid state (no solvent)  60  170 h  4 major, many minor - 36, 37 and 38  Cyclopentene  60  21.5 h  > 10 -38 and others  Cyclohexene  60  41 h  >7  - 36, 37 and others  4-Methylcyclohexene  70  40 h  >7  - 36, 39 and others  Substrate  Number of products - identity 4 - 36, 37, 38 and 39 3 - 36, 37, 38, also remaining 2 > 10 - 37, 38 and others 8 - 36, 37, 38, 39 and others  I  1 I  4.2.9.3 Reactivity of 2 with cyclic olefíns The reactivity of compound 2 with a variety of cycUc olefíns has been explored, and both the organic and organometallic products investigated. When 2 is thermolyzed in cyclopentene at 60 °C, more than ten peaks appear in the Cp región of the ' H N M R spectrum, indicating a multitude of organometallic products, one of which is identifíed as the known decomposition product 38. Small multiplets at 7.60 ppm, 0.39 ppm and -0.63 ppm suggest the presence of a metallacyclobutane complex (analogous to 25) in the mixture. No individual product can be isolated. The organic products are eluted from alumina with pentane in low yields. Lowresolution EIMS analysis indicates that oligomers of cyclopentene have formed. Thermolysis of 2 in cyclohexene leads to the formation of oligomers of cyclohexene. However, yields are low, with only 8 mg of oligomers obtained from a reaction that began with 100 mg of 2. A G C / M S analysis gives relative percentages of 19% neopentyl-capped dimer, 56% trimer, 6% neopentyl-capped trimer, and 15 % tetramer. No pentamers or higher oligomers are observed. A mixture of organometallic products is isolated by chromatography on alumina, and the corresponding ' H N M R spectrum has more than seven peaks in the Cp región. The largest of these singlets correspond to the chemical shift associated with decomposition products 36 and 37. Compound 2 in 4-methylcyclohexene under thermolysis conditions does not produce any oligomeric organic products. The two main organometallic products detected are 36 and 39, and there are also traces of 37 and other imidentified products in the ' H N M R spectrum. Thus, compound 2 shows some ability to oligomerize the simplest cyclic olefíns, but with littie effectiveness when compared to precatalyst 1. Decomposition pathways still predomínate over cyclic olefín oligomerization.  4.2.9.4 Reactivity of 2 with oxygen-containing cyclic olefíns Compound 2 has been thermolyzed in the three oxygen-containing cyclic olefins illustrated in Scheme 4.4 (vide suprd) in order to provide a comparison to 1. Thermolysis of 2 in 3,4-dihydro-2H-pyran at 70 °C does not yield any oligomeric organic products. A ' H N M R analysis of the organometallic products reveáis many peaks in the Cp región of the spectrum, the largest of which is diagnostic of 38. Compound 36 is also identified. A new product, unique to the reaction mixture, gives rise to ' H N M R resonances that include a Cp peak at 5.14 ppm, a strong Me región peak at 1.12 ppm, and several multiplets including two  clear doublets of doublets at 4.81 and 6.30 ppm. However, separation of this new compound from the other decomposition products has not worked well. Because of the lack of oligomerization, the reaction products have not been pursued further. A similar thermolysis of 2 in 2,3-dihydrofuran also yields a multitude of organometallic products, the main ones of which are identifíed as the decomposition products 36, 37 and 38. The ' H N M R spectra contain small multiplet peaks that might be similar to those observed in the 3,4-dihydro-2H-pyran product spectrum, but they carmot be analyzed or linked to any of the Cp peaks. No evidence for oligomeric products of 2,3-dihydrofiiran appears in the ' H N M R spectra. Thermolysis of 2 in 2,5-dihydroñiran produces a creamy precipítate in a light orange solution. Solvent extractions followed by ' H N M R analysis afford a spectrum with many broad and complex peaks. As such, interpretation of the spectrum is almost impossible. This lack of clear organometallic products from the reaction of 2 with 2,5-dihydroñiran is in sharp contrast to the formation of 29 as the main product from the reaction of 1 with the same substrate. Oligomeric products cannot be conclusively identified based on the ' H N M R data. This brief survey of the reactivity of 2 with the oxygen-containing cyclic olefins suggests that the decomposition pathways that lead to the formation of 36, 37 and 38 still domínate for these substrates. No evidence for oligomers of the oxygen-containing cyclic olefins has been found.  4.2.9.5 Reactivity of 2 with nitrogen-containing cyclic olefíns Compound 2 has also been reacted with 1,2,3,6-tetrahydro-pyridine. Within a few minutes of mixing at rt, the initially red solution becomes orange-red and a fine solid precipitates. Subsequent thermolysis at 70 °C yields a red-orange reaction mixture. ' H N M R analysis indicates many products, but no signáis attributable to Cp can be identified. This complex reactivity is in contrast to the quantitative formation of 33 from 1 and in contrast to the dominant decomposition pathways for thermolysis of 2. One G C / M S analysis contains a large peak at t = 12.6 min with an associated parent mass of 334 amu that is tantalizingly cióse to the predicted mass of a 1,2,3,6-tetrahydro-pyridine tetramer (332.5 amu). However, the result is not reproducible, ñor can the putative organic product be isolated. Therefore, it must be concluded that 2 does not cause the oligomerization of 1,2,3,6-tetrahydro-pyridine.  4.3 Summary Precatalyst 1 forms an alkylidene intermediate under thermolysis conditions and reacts with cyclic olefíns of various ring sizes. A cyclohexene-derived trans-metallacycle (24) is detected by ' H N M R spectroscopy, as is a cyclooctene trans-metallacycle (26), and both are comparable to their molybdenum analogues. Two cyclooctene units can couple in the tungsten's coordinatíon sphere to yield the 1,4-diene 27 and the allyl hydride 28, which represent a stage further along the catalytic reaction pathway presented in Chapter 5. Oligomerization proceeds best for cyclohexene, with cyclooctene forming isolable organometallic products that in tum show oligomerization reactivity. In contrast, cyclopentene and 2,5-dihydrofuran form stable cismetallacycles (25 and 29) that do not readily react ftuther. Introduction of a methyl group onto the cyclic substrate, as with 4-methylcyclohexene, lowers the yield of oligomers. Inclusión of an oxygen or nitrogen heteroatom in the substrate ring is also detrimental to oligomerization, as the preferred reaction then becomes C - 0 bond cleavage, N - H activation, or decomposition. Thus, the oligomerization ability of precatalyst 1 seems limited to simple cyclic olefíns of appropriate size. The isolable products give insight into the reaction mechanism, which will be explored next in Chapter 5. The reactivity of 2 has been explored with various substrates, including the cyclic olefíns. As with 1, an alkylidene intermediate of 2 can be trapped with PMea under thermolysis conditions. However, unlike compound 1, compound 2 does not cleanly actívate C-H bonds, forming instead a number of solvent-independent, dimeric decomposition products. The reactivity of 2 with representative cyclic olefíns shows that the complex has some limited ability to oligomerize the substrates, but no fijrther insights on the mechanistic aspects of this chemistry are revealed.  4.4 Experimental Procedures  4.4.1 General Methods A l l reactions and subsequent manipulations involving organometallic reagents were performed imder anaerobic and anhydrous conditions either at a vacuum-nitrogen dual manifold or in an inert-atmosphere dry box. Pentane, benzene-í/ó, diethyl ether, and tetrahydrofuran (THF) were all dried over sodium benzophenone ketyl and were freshly distilled prior to use. Cyclopentene, cyclohexene, and cyclooctene were all purchased from Aldrich, and were dried over sodium benzophenone ketyl, distilled, and stored in resealable glass vessels. 2,5dihydrofliran was dried twice with sodium, once with sodium benzophenone ketyl, distilled, and stored in a resealable glass vessel. Deoxygenated 2,3-dihydro-2H-pyran was shaken over alumina and fíltered through a glass-wool plug before use. 1,2,3,6-tetrahydro-pyridine was vacuum transferred from calcium hydride before use. Cp*W(NO)(CH2CMe3)2 (I),'* was prepared according to the published procedures. CpW(NO)(CH2CMe3)2 (2) was not prepared according to the original method^^ but according to a modified versión of the Cp*Mo(NO)(CH2CMe3)2 preparatíon.^'' A l l other chemicals were purchased from Aldrich and were used as received. A l l IR samples were prepared as Nujol muUs sandwiched between NaCl plates, and their spectra were recorded on a Thermo Nicolet Model 4700 FT-IR spectrometer. N M R spectra were recorded at room temperature on Bruker AV-300, Bruker AV-400 or Bruker AV-500 instruments using standard U B C pulse sequences with delay tímes of 1 sec for ID experiments and 1.5 sec for 2D experiments. A l l chemical shifts and coupling constants are reported in ppm and in Hertz, respectively. ' H N M R spectra were referenced to the residual protio isotopomer present in CgDé (7.15 ppm). '^C N M R spectra were referenced to CeDe (128 ppm). Where necessary, ' H - ' H C O S Y , ' H - ' ^ C H M B C , ' H - " C H M Q C , and '^C A P T experiments were carried out to correlate and assign ' H and '^C N M R signáis. GC/MS analyses were carried out on an Agilent 6890 Series G C system equipped with a non-polar, cross-linked 5% diphenyl- 95% dimethylpolysiloxane column and an Agilent 5973 Network mass-selective detector. Lowresolution mass spectra (El, 70 eV) were recorded by the staff of the U B C mass spectrometry facility using a Kratos MS-50 spectrometer. Elemental analyses were performed by Mr. Minaz Lakha of the U B C microanalytical facility.  4.4.2 Reactivity of 1 in cyclohexene: Identification of 24 Reaction conditions for the thermolysis of 1 in cyclohexene and characterization data for the oligomeric products are presented in Chapter 2. Isolation of the organometallic products was attempted via chromatography on alumina columns with pentane/EtiO mixtures of various rallos. No individual organometallic products were isolated or ñiUy identified. Selected resonances for 24 obtained from a 7:1 pentane:Et20 fraction: ' H N M R (CeDfi, 500 MHz, 25 °C) 6 -1.07 (IH, m, Hl), 0.09 ( I H , dq, VHH = 4.8, VHH = 13.1, H2), 1.22 (9H, s, CMe^), 1.74 (15H, s, CsMes), 3.64 (IH, d, VHH = 4.6,//7). In many cases, EtaO fractions also contain strong singlets at 1.35 and 1.53 ppm, even when 24 is absent. This product(s) has not been identified.  4.4.3 Reactivity of 1 in cyclopentene: Preparation of 25 A red 0.02 M solution of 1 (0.050 g, 0.102 mmol) in cyclopentene (5.0 mL) was thermolysed at 70 °C for 49.5 h, and the final reaction mixture was orange-red. The volátiles were removed in vacuo and the reaction residue (0.093 g) was analyzed by ' H N M R spectroscopy. The solvent was removed in vacuo, and the reaction residue was redissolved in EtaO and placed in a -32 °C freezer ovemight. A n orange precipítate was isolated and identified as the 2 + 2 metallacycle 25 (0.038 g, 0.078 mmol, 76% yield). The remaining reaction products were chromatographed on an alumina column (3 x 0.5 cm). Colorless oil (0.039 g) was recovered from the pentane fraction, and a bronze-brown residue (0.018 g) was recovered from the EtiO fraction. 'HNMR  (CeDfi, 300 M H z , 25 °C) 6 -0.48 ( I H , m, Hl), 0.16 (IH, m,  CH2), 1.17 (9H, s, CMes), 1.53 (IH, m, CH2), 1.74 (15H, s, CsMcj), 2.07  4  (IH, m, CH2), 2.21 (IH, m, CH2), 2.45 (IH, m, CH2), 3.18 (IH, d, VHH = 4.9, H6), 7.65 (IH, dt, VHH = 8.2, VHH =11-4, H2). One CH2 protón signal is obscured, presumably under one of the strong methyl singlets.  ' ^ C { ' H } N M R (CéDe, 125 M H z , 25 °C) 6 10.2 (CjMes), 15.9 (Cl), 32.5 {CMe^), 33.7 (C3, C4 or C5), 37.6 (C3, C4 or C5), 40.3 (C3, C4 or C5), 41.6 (CMes), 108.9 (CgMes), 115.0 (C2 or C6), 139.9 (C2 or C6). M S (El, 120 °C): miz 487 [P*].  Isolated oligomeric organic products. MS (El, 150 °C  300 °C): miz 135, 201, 269,  337, 405, 473, 540, 608, 676, 744 [P^]. GC/MS (standard method) dimers 0.5 min, miz = 136; trimers 4-7.5 min, miz = 204, 202; tetramers 10-12 min, miz = 272, 270; pentamers 13-14 min, miz = 340, 338; hexamers 16-17 min, miz = 406, 404; septamers 18-19 min, miz = 474, 472. The bronze-brown residue contained many other organometallic decomposition products which were not identified ( ' H N M R spectrsocopy).  4.4.4 Reactivity of 1 in cyclooctene: Preparation of 27 and 28, identifícation of 26 A red 0.01 M solution of 1 (0.203 g, 0.413 mmol) in cyclooctene (41.0 mL) was thermolysed at 70 °C for 48 h. The volátiles were removed in vacuo, and the reaction residue (0.321 g) was analyzed by ' H N M R spectroscopy. Complex 26 was identified on the basis of characteristic N M R signáis in the spectra. The reaction residue was chromatographed on an alumina column (3.5 x 2 cm). Clear, slightly yellowish oil (0.066 g) was recovered fi-om the pentane fraction, and an orange-red residue (0.090 g) was recovered from the EtaO fraction. The orange-red residue was selectively extracted with pentane to obtain a red solid (0.056 g, 0.099 mmol, 24%) and an orange solution. The red solid was dissolved in THF and pentane was diffused into the solution at rt to form small red crystals of 27. Yellow crystals of 28 deposited together with red crystals of 27 from the orange solution upon cooling, and they were separated by hand. Compound 26: selected ' H N M R resonances (CeDe, 300 MHz, 25 °C) 5  ,..wO  -0.49 (IH, dq, VHH =11.6, VHH = 3.4, Hl), 1.69 (15H, s, CsMcj), 3.98 (IH, dt, VHH = 10.0, VHH = 4.5, H2), 4.03 (IH, d, VHH = 4.2, H9).  Compound 27: IR (Nujol): VNO 1577 (s) cm"'. ' H N M R (CeDe, 400 M H z , 25 °C) 5 -1.26 (IH, pseudo t, VHH = 10.2, H9), 0.90 (IH, ddd, VHH = 6.3, VHH = 9.9, VHH = 11.3,//7i), 1.19 (IH, m, CH2), 1.38 (4H, m, CH2), 1.58 (3H, m (obscured), CH2), 1.65 (15H, s, CjMes), 1.71 (3H, m (obscured), CH2), 1.84 (3H, m, CH2), 1.92 (3H, m, CH2), 2.14 (IH, dd, VHH = 2.8, VHH = 10.5, Hl), 2.30 (3H, m, CH2), 2.45 (2H, m, CH2), 4.48 ( I H , m, HIO). ' ^ C I ' H } N M R (CeDe, 100 M H z , 25 °C) 5 9.5 (C9), 10.1 {CsMes), 26.6 (CH2), 26.7 (CH2), 27.5 (CH2), 27.8 (CH2), 28.2 (CH2), 28.9 (CH2), 30.7 (CH2), 31.7 (CH2), 31.9  (CHz), 32.7 (CH2), 34.6 (CH2), 42.3 (CIO), 47.5 ( C l l ) , 69.4 (C8), 72.1 ( C l ) , 102.1 (CsMes). HRMS-EI miz: [M^] caled for '^^WNOC26H4i 567.26977; found 567.26968. Compound 28: IR (Nujol): VNO 1589 (s) cm''. ' H N M R (CgDé, 400 MHz, 25 °C) 5 0.12 (IH, s with satellites, 'JWH = 124.6, WH), 0.54 (IH, dd, VHH = 14.7, VHH = 6.7, CH2), 1.32 (7H, overlapping m, CH2), 1.55 (5H, overlapping m, Hl and CH2), 1.76 (15H, s, CsMcs), 1.93 (3H, overlapping m, CH2), 2.03 ( I H , m, CH2), 2.22 (IH, m, CH2), 2.44 (2H, overlapping m, CH2), 2.92 (3H, overlapping m, CH2), 5.17 (IH, ddd, VHH = 12.4, VHH = 9.6, VHH = 7.3, HU),  7.00 (IH, d, VHH = 12.4, HIO). ' ^ C { ' H } N M R  (CeDe, 100 M H z , 25 °C) 5 10.4 (CsMes), 20.9 (CH2), 26.0 (CH2), 26.7 (CH2), 27.5 (CH2), 28.4 (CH2), 28.7 (CH2), 29.7 (CH2), 31.0 (CH2), 31.7 (CH2), 32.1 (CH2), 35.0 (CH2), 64.0 ( C l ) , 91.2  (C9), 105.3 (CsMes), 119.1 ( C l l ) , 120.8 (C8), 138.8 (CIO). M S (El, 100 °C): miz 567 [P^]. Data for organic products: ' H N M R (CeDe, 400 M H z , 25 °C) 6 0.85-0.95, 1.2-1.75, 1.902.50, 3.10-3.25, 5.2-5.4, 5.4-5.55, 5.6-5.8; aliphatic to olefinic protón ratio 9:1. M S (El, 150 °C ^ 300 °C): miz 218, 328. M S (El, 150 °C): miz 441, 551, 661, 771. G C / M S (standard method): cyclooctene capped with neopentyl, one unsaturation {miz = 180) at 1.4 min; cyclooctene capped with neopentyl, two unsaturations {miz = 178) at 2.9-3.2 min; cyclooctene dimers, two unsaturations {miz = 218) at 7.2-8.5 min; cyclooctene dimer capped with neopentyl {miz = 288) at 12-12.5 min; cyclooctene trimer, two unsaturations {miz = 328) at 13.2-13.8 min.  4.4.5 Reactivity of 1 in 4-methylcycIohexene A red 0.16 M solution of 1 (0.1006 g, 0.205 mmol) in 4-methylcyclohexene (1.3 mL) was thermolysed at 70 °C for 40 h, and the final reaction mixture was dark brown. The volátiles were removed in vacuo, and the reaction residue was extracted with pentane. The pentane was removed, and the product mixture was analyzed by ' H N M R spectroscopy. The reaction residue was then chromatographed on an alumina column (2 x 0.5 cm). The pentane fraction contained the oligomeric organic products (8 mg). The Et20 fraction contained the organometallic products (71 mg). The isolated oligomeric products: ' H N M R (CeDg, 500 M H z , 25 °C) 5 0.7-1.0,1.1-1.2, 1.4-2.2 (aliphatic protons), 5.2-5.8 (olefinic protons); all were broad, complex peaks. GC/MS analysis (standard method) peaks: dimers did not appear, dimers capped with a neopentyl unit at  13-1.6 min (4 peaks, 9%, miz = 262), trimers at 9.9-11.3 min (many peaks in a broad lump, 72%, miz = 286), trimers capped with a neopentyl unit at 12.4-13.3 min (6-7 peaks, 3%), miz = 356), tetramers at 14.1-14.9 min (-15 peaks in a broad lump, 16%, miz = 380, 382). The baseline rose between 17 and 18 min, but this could not be assigned to the presence of pentamers conclusively. The organometallic products: a ' H N M R spectrum of the crude product mixture had many peaks in the methyl región. No main product could be identifíed. A ' H N M R spectrum of the contents of the EtaO fraction eluted from alumina also had many methyl-region peaks. The presence of the decomposition oxo product Cp*W(0)2(CH2CMe3) was indicated by the characteristic peaks at 1.32 and 1.66 ppm. No other organometallic products could be isolated or identified. A blank of 4-methylcyclohexene was also thermolysed imder identical experimental conditions in the absence of 1. E I S M analyses (100 °C) of the blank and of the isolated reaction volátiles both showed only 4-methylcyclohexene (miz = 96).  4.4.6 Reactivity of 1 in 2,5-dihydrofuran: Preparation of 29 A red 0.11 M solution of 1 (0.470 g, 0.956 mmol) in 2,5-dihydrofijran (8.5 mL) was thermolysed at 70 °C for 40 h, and the final reaction mixture was dark. The volátiles were removed in vacuo. The reaction residue was extracted with pentane and the extracts were filtered through a celite column (2 x 0.5 cm) to obtain a dark solution. The remaining orange residue was then extracted with Et20 and the extracts fihered through the celite column to obtain an orange solution. The orange main product 29 was isolated in 48% yield (0.224 g, 0.458 mmol) in múltiple crops. Crystals of 29 suitable for X-ray crystallography were grown from Et20. Complex 29 did not elute firom an alumina column. The dark-colored minor products could not be identified. No organic products were observed or recovered. Anal. Caled for C,9H3iWN02: C, 46.64; H , 6.39; N , 2.86. Found: C, 47.00; H , 6.39; N , 3.26. IR (CCI4): VNQ 1560 (s) cm"', (Nujol): VNO 1544 (s) cm-'. ' H N M R (CeDg, 500 M H z , 25 °C) 5 -0.46 (IH, m, H3), 1.07 (9H, s, CMci), 1.74 (15H, s, CsMes), 3.13 (IH, dd, VHH = 8.75, VHH = 4.23, H4a), 3.61 (IH, ddd, VHH = 9.11, VHH = 4.58, VHH = 1-29, Hla), 3.75 (IH, d, VHH = 4.49, H5), 3.77 (IH, dd, VHH = 8.79, VHH = 5.44, H4b), 4.49 (IH, dd, VHH =  9.14, VHH = 6.91, Hlb), 6.67 ( I H , m, H2). ' ^ C { ' H } N M R (CeDg, 125 M H z , 25 °C) 5 10.1  (CsMes), 15.1 (C5), 32.3 (CMes), 42.9 (Cd), 72.7 (two overlapping peaks, Cl and C4), 109.4 (CsMes), 117.3 (C2), 128.9 (C5). M S (El, 100 °C): m/z 489 [P^].  4.4.7 Reactivity of 1 in 2,3-dihydrofuran A red 0.03 M solution of 1 (0.149 g, 0.303 mmol) in 2,3-dihydrofuran (10 mL) was thermolysed at 70 °C for 44 h, and the final reaction mixture was dark. The volátiles were removed in vacuo. The dark, oily reaction residue was extracted with pentane to obtain dark products (0.186 g). The remaining residue was extracted with EtiO with similar results. ' H N M R analysis revealed the presence of many products, none of which could be isolated or identified. No organic products were observed or recovered.  4.4.8 Reactivity of 1 in 3,4-dihydro-2H-pyran: Preparation of 30 A red 0.02 M solufion of 1 (0.074 g, 0.151 mmol) in 3,4-dihydro-2H-pyran (8.0 mL) was thermolysed at 70 °C for 44 h, and the final reaction mixture was dark. The volátiles were removed in vacuo to obtain a dark red, oily residue (0.101 g). Chromatography on an alumina column (2 x 0.5 cm) eluting with pentane:Et20 yielded a red-brown residue (0.024 g) and then with THF yielded a dark brown residue (0.021 g). The pentaneiEtaO residue was selectively extracted with pentane, and then the remaining residue was extracted with Et20 to isolate the product of interest. In Et20, 30 precipitated as small light tan clusters of microcrystals. IR (Nujol): VNO 1562 (s) cm''. ' H N M R (CéDe, 300 MHz, 25 °C) 6 I. 18 (9H, s, CMes), 1.41 (IH, m, H2a), 1.67 (15H, s, CsMcs), 1.97 (3H, overlapping m, H2b and both H6), 2.34 (2H, overlapping dd Q W 3 4 1  2  and m, VHH = 9.9, VHH = 14.9, H3 and H5), 3.52 (IH, dt, VHH = II. 7, VHH = 5.0, Hla), 3.95 (IH, ddd, VHH =11.8, VHH = 8.4, VHH = 3.7, Hlb), 4.35 (IH, ddd, ^JHH = 9.9, VHH = 3.8, VHH = 0.9, H4).  ' ^ C { ' H } N M R (CeDó, 75 M H z , 25 °C) 5 10.3 (CsMes), 29.3 (C2), 30.5 (CMes), 30.8 (CMcs), 32.1 (C6), 43.5 (C5), 59.2 (C3), 62.9 (Cl), 81.5 (C4), 117.0 (CsMes).  4.4.9 Reactivity of 5 witli oxygen-containing heterocycles: Preparation of 31 and 32 Chris Semiao synthesized 31 by thermolysing 5 in 2,3-dihydro-2/f-pyran at 45 °C for 24 h. The volátiles were removed in vacuo, and the residual solid was washed with EtaO and dissolved in THF. X-ray quality crystals of 31 were obtained by evaporation of the solvent at rt, followed by washing with cold THF and EtaO to sepárate the crystals from the matrix. Anal. Caled for CaaHasNOW: C, 51.70; H , 5.28; N , 2.62. 7 8/ 0^^^"^"^^^/56 ^\ 1  \\  ^""^^^  FoundC,51.80;H, 5.19; N , 2.89. IR (Nujol): VNO = 1556 "^^^  ^ 7.51-7.01 múltiple  / 3 4  signáis (6H, m overlapping, Haryi and H6), 6.14 (IH, d  2  n v e r l a n n i n a 3/uu 16 H7), HT\ 6.11 6 11 (IH, í'IH dd rlrl overlapping, nv^rlannino VHH ^, = overlapping, ^HH == 16,  13, VHH = 9, H4), 5.30 ( I H , ddd, VHH = 9, VHH = 6, VHH = 3, Hla), 3.92 (IH, td, VHH = 9, VH 7HH = 9, VHH = 4, HJb), 3.73 (IH, ddd, VHH = 6, VHH = 10, VHH = 13, H3), 2.24 (IH, t, VHH = 9, H5), 1.62 (15H, s, CsMes), 1.60-1.53 múltiple signáis (2H, m overlapping, both H2). " C N M R (CéDé, 100 M H z , 25 °C) 6 139.0 (C8), 133.5 (C6), 129.4 (Caryi), 127.1 (Caryi), 126.7 (Ca,y,), 124.3 (C7), 116.1 (C4), 111.7 (C3), 109.5 (CsMes), 90.5 (Cl), 61.8 (C5), 35.3 (C2), 9.8 (CsMes). M S (El, 100 °C): m/z 535 [P^]. lan Blackmore synthesized 32 in an analogous manner by thermolysis of 5 in 2,5dihydrofuran.  4.4.10 Reactivity o f l in 1,2,3,6-tetrahydro-pyridine: Preparation of 33 In a J. Young N M R tube with a CeDe reference capillary, a red 0.09 M solution of 1 (0.042 g, 0.085 mmol) in 1,2,3,6-tetrahydro-pyridine (1 mL) was thermolysed at 70 °C for 62 h. The final reaction mixture was bright orange. The volátiles were removed in vacuo to obtain a dark orange residue (0.054 g). The reaction residue was extracted twice with EtaO to obtain an orange solution, which was stripped of solvent to obtain a crude orange solid (0.042 g). A white precipítate (0.003 g) remained after the extraction. Recrystallization of the orange solid in EtaO at -30 °C isolated 33 as bright orange crystals (0.022 g in two crops, 0.044 mmol, 51% yield). Chromatography of the mother liquor on an alumina column (1.5x0.5 cm) isolated no organic products. A sample of 33 heated in CeDe at 100 °C for one week darkened slightly and ' H N M R analysis showed large peaks for 33 and only traces of decomposition.  IR (Nujol): VNO 1567 (s) cm''. M S (El, 100 °C): m/z 502 [P""]. Initial mp = 142 °C, remelt mp = 135 °C. Major isomer of 33, ' H N M R (CeDe, 300 M H z , 25 °C) 5 0.97 (2H, d, VHH = 3.4, H6), 1.40 (9H, s, CMes), 1.65 (15H, s, CsMes), 1.77 (IH, obscured, H4), 2.00 (IH, br m, H4), 3.44 (2H, br m, both Hl), 3.90 (IH, dt, VHH = 1 'N {  11.1 Hz, VHH = 4.3 Hz, H5), 4.98 (IH, br dd with fine structure, H5),  T 3  5.23 ( I H , br d with fine structure, H2), 5.51 (IH, br m with fine structure, H3). ' ^ C { ' H } N M R (CeDe, 75 M H z , 25 °C) 5 9.7 (CsMes), 28.7 (C4), 34.6 (CMes), 37.8 (C7), 57.4 (C(5), 59.5 (Cl), 65.2 (C5), 109.8 (CsMes),  126.0 (C2), 127.0 (C3). Minor isomer of 33, selected peaks, ' H N M R (CeDe, 300 M H z , 25 °C) 5 1.04 (2H, d, VHH = 4.8, C/ZiCMes), 1.41 (9H, s, CMca), 1.64 (15H, s, CsMes), two protón signáis obscured (NCH2C//2), 3.17 (2H, m, NC//2CH=CH), 4.30 (IH, br d, VHH = 15.3, NC//2CH2), 5.23 (IH, obscured, N C H 2 C / / = C H ) , 5.37 (IH, br d with fine structure, NC//2CH2), 5.51 (IH, obscured, N C H 2 C H = C i í ) . Selected peaks " C { ' H } N M R (CeDe, 75 M H z , 25 °C) 6 9.7 (CsMes), 25.9, 34.7 (CMes), 37.0, 56.8 (CH2CMe3), 123.3. The white powder: M S (El, 150-350 °C): m/z 351, 407, 502 [P^].  4.4.11 Reactivity of 1 in 3-pyrroline In a J. Young N M R tube with a CeDe reference capillary, a red 0.1 M solution of 1 (0.045 g, 0.092 mmol) in 3-pyrroline (0.9 mL) was analyzed by ' H N M R spectroscopy and then thermolyzed at 70 °C for 17 h. The final reaction solution was palé yellow, and ' H N M R analysis in situ showed the consumption of 1 and the formation of new products. The volátiles were removed in vacuo to obtain a yellow-orange residue (0.036 g). The reaction residue was dissolved in CeDe and analyzed. Selective solvent extractions and chromatography failed to sepárate any single product. The complex ' H N M R spectra could not be interpreted. The addition of Et20 to the reaction residue caused a white powdery precipítate to form. No organic oligomers were detected. The white powder: M S (El, 150-350 °C): m/z 271, 287, 343, 407 [P^].  4.4.12 Reactivity of 2 with trimethylphosphine in THF and in cyclohexene Trimethylphosphine (0.5 mL) was vacuum transferred onto a frozen 0.02 M solution of 2 (0.0774 g, 0.184 mmol) in THF (10 mL). The reaction mixture was brought to rt and then  thermolyzed at 60 °C for 40.5 h. The volátiles were removed in vacuo, and the orange residue was dissolved in EtaO and filtered through a celite column (2 x 0.5 cm). The volume of the solution was reduced, and it was placed in a -30 °C freezer to deposit an orange crystalline precipítate. The ' H N M R spectrum of the product showed the characteristic peaks arising from 34.3 ^ similar thermolysis of 2 in cyclohexene with trimethylphosphine afforded only 34, as detected by ' H N M R spectroscopy. Selected ' H N M R resonances of CpW(NO)(CHCMe3)(PMe3) (34): (CeDg, 500 M H z , 25 °C) 5 1.07 (9H, dd, VHP = 9.6, VHH = 0.9, PMe3), 1.43 (9H, s, CMe3), 5.24 (5H, s, C5H5), 12.01 (IH, d, 3JHP = 3.0, alkylidene H).  4.4.13 Reactivity of 2 with trimethylphosphine in cyclohexane Trimethylphosphine (0.25 mL, -2.5%) was vacuum transferred onto a frozen 0.02 M solution of 2 (0.0773 g, 0.184 mmol) in cyclohexane (10 mL). The reaction mixture was brought to rt and then thermolysed at 60 °C for 25 h. The volátiles were removed in vacuo, and the dark reaction residue was dissolved in CeDe. ' H N M R analysis showed the characteristic peaks arising from 34.3 Trimethylphosphine (0.2 mL, -1.3%) was vacuum transferred onto a frozen 0.008 M solution of 2 (0.0480 g, 0.114 mmol) in cyclohexane (15 mL). The reaction mixture was brought to rt and then thermolysed at 60 °C for 19 h. The volátiles were removed in vacuo, and the dark red-orange reaction residue was dissolved in CeDe. ' H N M R analysis showed the characteristic peaks arising from remaining 2, from 34, and from 35.3  relative ratio of the three  compounds, based on integration of the Cp peaks, was 4:2:1. Selected ' H N M R resonances of CpW(NO)(CeHio)(PMe3) (35): (CeDe, 500 MHz, 25 °C) 6 1.13 (9H, d, VHP = 8.8, PMe3), 4.87 (5H, s, C5H5).  4.4.14 Reactivity of 2 in cyclohexane: Formation of 36,37,38,39 and 40 A red 0.03 M solution of 2 (0.2890 g, 0.686 mmol) in cyclohexane (25 mL) was thermolysed at 60 °C for 41 h, and the final reaction mixture was dark. The volátiles were removed in vacuo. The reaction residue was dissolved in pentane and fíltered through a celite column (2 x 2.5 cm). The four major organometallic products detected by ' H N M R spectroscopy were named 36,37,38 and 39.  Numerous attempts to sepárate the products by selective precipitation and the use of column chromatography were partially successful. Complex 39 and other minor products did not elute from a Florisil column, but 36,37 and 38 were recovered. Then when an alumina column was employed, product 37 was isolated from the EtaO fraction, product 36 was isolated from the THF fraction, and product 38 did not elute from the column. When the reaction was performed at higher concentrations (0.05M) a fifth product, 40, was isolated with 36 in the THF fraction chromatographed on an alumina column. In addition, i f a solution of 2 in cyclohexane (0.03M) was left at rt for 34 days, ' H N M R analysis revealed the presence of equal amounts of 36, 37 and 38, as well as remaining 2. Complex 36, W2N2O2C30H54: selected ' H N M R resonances (CeDé, 500 M H z , 25 °C) 6 1.02 (9H, s, methyl), 1.25 (9H, s, methyl), 1.97 (IH, d, VHH = 13.4), 2.42 (IH, d, VHH = 13.4), 3.94 (2H, s), 5.61 (5H, s, C5H5). M S (El, 150 °C): miz 685 [P - 2(CH2CMe3) - Me]. Complex 37, W2N2O2C30H52: selected ' H N M R resonances (CeDe, 500 MHz, 25 °C) 6 1.22 (9H, s, methyl), 1.43 (9H, s, methyl), 2.04 (IH, d, VHH = 13.9), 2.28 (IH, d, VHH = 14.0),  5.28 (5H, s, C5H5), 9.92 (IH, s). M S (El, 150 °C) miz 11\ [P - C5H9]. Complex 38, W2N2O2C20H32: selected ' H N M R resonances (CeDé, 500 M H z , 25 °C) 5 1.26 (9H, s, methyl), 2.12 (2H, s), 5.42 (5H, s, C5H5). Complex 39: selected ' H N M R resonances (CÓDÓ, 500 M H z , 25 °C) 5 0.73 (9H, s, methyl), 1.38 (9H, s, methyl), 5.23 (lOH, s, two C5H5). Complex 40: selected ' H N M R resonances (CeDg, 500 M H z , 25 °C) 6 0.78 (9H, s, methyl), 1.61 (9H, s, methyl), 2.73 (IH, d, VHH = 16.2), 2.97 ( I H , d, VHH = 15.9), 5.40 (5H, s,  C5H5), 5.50 (5H, s, C5H5). 4.4.15 Thermolysis of 36 and 37 in CeDe A solution of 36 and 37 in CeDe was heated at 60 °C, and monitored by ' H N M R spectroscopy. After 2.75 h, signáis attributable to 36, 37 and 38 in a 2:1:1 ratio were observed. Añer 27 h the ratio had changed to 2:1:2. After 71 h signáis for 36 had disappeared, giving a ratio of 0:1:4. Finally, after 93 h the ratio was 0:1:7.5. As the signáis due to product 36 disappeared and those for 38 grew in, neopentane was detected in the product mixture as evidenced by a singlet at 0.80 ppm.  4.4.16 Thermolysis of 2 in a range of other substrates Complex 2 was thermolyzed at reaction temperatures ranging from 40 °C to 70 °C in a range of substrates including tetramethylsilane, mesitylene, dichloromethane, cyclopentene, cyclohexene, 4-methylcyclohexene, 3,4-dihydro-2H-pyran, 2,3-dihydrofuran, 2,5-dihydrofuran, and 1,2,3,6-tetrahydro-pyridine. Each reaction produced a plethora of products, often including 36, 37, and 38, as summarized in Table 4.2 (vide suprd). The reaction in cyclohexene is described as a representative example. A red 0.07 M solution of 2 (0.089 g, 0.211 mmol) in cyclohexene (3 mL) was thermolysed at 60 °C for 41 h, and the final reaction mixture was dark. The volátiles were removed in vacuo. The reaction residue was chromatographed on an alumina column (2 x 0.5 cm). Organometallic products were recovered from the EtiO fraction. The ' H N M R spectra of the organometallic products in CeDe contained more than seven peaks in the Cp región, the largest of which could be identified as arising from 36 and 37. The organometallic products were not further identified. Thermolysis of 2 in 3,4-dihydro-2H-pyran produced many products including 36 and 38, and one new product. The selected ' H N M R resonances for that unidentified product follow: ' H N M R (CeDe, 500 M H z , 25 °C) 6 1.12 (s. Me), 2.97 (m), 3.74 (t, J = 11.0), 4.25 (m), 4.81 (dd, J = 1.7, J = 16.1), 5.14 (s, Cp), 6.30 (dd, J= 5.5, J = 16.1).  4.4.17 Thermolysis of 2 in the solid state Sohd 2 (0.0703 g, 0.167 mmol) was thermolysed at 60 °C for 170 h. The black residue was dissolved in pentane and filtered through a celite column. The ' H N M R spectrum of the products in CeDe contained many peaks in the Cp región. Three of the largest ones were characteristic of 36, 37 and 39. Some starting material also remained, as evidenced by its characteristic Cp peak. The decomposition products were not further identified.  4.4.18 X-ray crystallography Data collection for each compound was carried out at -100 ±1 °C on a Rigaku AFC7/ADSC C C D diffiactometer or on a Bruker X8 A P E X diffractometer, using graphitemonochromated Mo K a radiation.  Data for 27 were coUected to a máximum 16 valué of 56.2 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. Disorder at C14 and C15 was modeled as Part A (70%) and Part B (30%). As a result, C14a, C14b, C15a and C15b were refined isotropically, and all other non-hydrogen atoms were refined anisotropically. Hydrogen atoms H l , HIO, and H l 1 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 5172 observed reflections and 277 variable parameters. Data for 28 were coUected to a máximum 26 valué of 66.8 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. The disordered Cp* ring was modeled in three co-planar orientations; the largest contributor was modeled anisotropically and the two lesser contributers were modeled isotropically. A l l other nonhydrogen atoms were refined anisotropically. Hydrogen atoms H l , HIO, and H l 1 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The hydride on the tungsten center could not be modeled. The final cycle of fiíU-matrix least-squares analysis was based on 9241 observed reflections and 271 variable parameters. Data for 29 were coUected to a máximum 26 valué of 56.0 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 4487 observed reflections and 216 variable parameters. Data for 31 were coUected to a máximum 26 valué of 56.0 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H3, H4, H5, H6 and H7 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 4876 observed reflections and 269 variable parameters. Data for 32 were coUected to a máximum 26 valué of 56.4 ° in 0.5 ° osciUations. The structure was solved by direct methods^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H2, H3 and H5 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of  fiíll-matrix least-squares analysis was based on 4525 observed reflections and 252 variable parameters. Data for 33 were collected to a máximum 16 valué of 56.2 ° in 0.5 ° oscillations. The structure was solved by direct methods^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H5a and H5b were refined isotropically, and all other hydrogen atoms were included in fixed positions. The double bond was set at C2-C3; the bond lengths indícate that the actual position is averaged between C2-C3 and C3-C4. The final cycle of full-matrix least-squares analysis was based on 4705 observed reflections and 233 variable parameters. For each structure neutral-atom scattering factors were taken from Cromer and Waber.^ Anomalous dispersión effects were included in  Fcaic;^  the valúes for A / and A / ' were those of  Creagh and McAuley.^ The valúes for mass attenuation coefficients are those of Creagh and H u b b e l l . A l l calculations were performed using the CrystalClear software package of Rigaku/MSC," or Shelxl-97.'' X-ray crystallographic data for the six structures are presented in Table 4.3, and in the cif files.  Table 4.3 X-ray Crystallographic Data for Complexes 27, 28, and 29. 27  28  29  Empirical fonimla  C26H41NOW  C26H4,NOW  C19H31NO2W  Crystal Habit, color  Prism, red  Prism, yellow  Prism, orange  Crystal size (mm)  0.10x0.07x0.05  0.45 X 0.40 X 0.35 0.2 X 0.2 X 0.2  Crystal system  Triclinic  Monoclinic  Monoclinic  Space group  P-i  P2i/a  P2i/n  Volume (Á^)  1154.6(3)  2393.14(14)  1875.70(6)  a (A)  9.7773(15)  14.6050(5)  8.1877(1)  biA)  11.3376(17)  8.6048(3)  15.1554(3)  c(A)  11.7186(18)  20.1135(7)  15.3309(3)  an  77.630(6)  90  90  pn rn  65.729(5)  108.781(1)  99.606(1)  88.019(6)  90  90  z  2  4  4  Density (calculated) (Mg/m^)  1.632  1.575  1.733  Absorption coefficient (mm"')  5.019  4.843  6.168  Fboo  572  1144  968  Measured Reflections: Total  17232  36804  30721  Measured Reflections: Unique  5172  9241  4487  Final R índices"  R l = 0.0677, wR2  R l = 0.0352, wR2  R l = 0.0161, w  = 0.1773  = 0.0864  = 0.0430  1.044  1.054  1.155  2.933 and-1.684  0.378 and-1.09  Crystal Data  Data CoUection and Refinement  Goodness-of-fit on  *  Largest diff peak and hole (e^ 4.694 and -4.014  " R l o n F = Z 1 (|Fo| - |Fc|) | / S |Fo|, ( / > 2a(7)); wR2 = [ (2 (F,^ -F.'f)/! data); w = [ a^o^ ]"'; * GOF = [ E (w (|Fo| - |Fc| )^ ) / degrees offreedom  w(Fo' ff' .  (all  Table 4.3 X-ray Crystallographic Data for Complexes 31,32 and 33. 31  32  33  Empirical formula  C23H29NO2W  C22H27NO2W  C20H34N2OW  Crystal Habit, color  Prism, orange  Prism, orange  Block, orange  Crystal size (mm)  0.63 X 0.25 X 0.20 0.30 X 0.22 X 0.07 0.5 X 0.4 X 0.2  Crystal system  Monoclinic  Monoclinic  Space group  P2i/c  P2i/n  Volume (Á^)  2046.2(5)  1863.0(5)  991.5(2)  a (A)  8.4925(14)  14.671(2)  8.2309(10)  6 (A)  11.8334(16)  9.4452(16)  9.3822(12)  c(A)  20.376(3)  14.673(2)  14.3092(19)  an  90  90  91.360(7)  92.221(6)  113.616(7)  96.015(6)  90  90  115.192(6)  z  4  4  2  Density (calculated) (Mg/m^)  1.738  1.859  1.683  Absorption coefficient (mm'')  5.663  6.217  5.834  1056  1024  500  Measured Reflections: Total  28872  15343  20888  Measured Reflections: Unique  4876  4525  4705  Final R índices''  R l = 0.0166, wR2 R l = 0.0247, wR2 R l = 0.0256, wR2  Crystal Data  Triclinic  Data Collection and Refínement  Goodness-of-fit on  *  Largest diff. peak and hole (e"  = 0.0407  = 0.0500  = 0.0667  1.097  1.111  1.165  1.173 and -0.943  1.697 and-0.846  4.495 and -2.093  A-^) R l on F = i : I (|Fo| - |Fe|) | / E |Foi, (/> 2^(7)); wR2 = [ (E ( Fo' - Fe' ) ' ) / S w(Fo' ff^ data); w = [ CT'FO' ]"'; * GOF = [ I (w ( |Foi - |Fe| f ) / degrees offreedom  .  (all  4.5 References (1)  a) Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, 223-233 and references cited therein. b) Blackmore, I. J.; Jin, X . ; Legzdins, P. Organometallics 2005, 24, 4088-4098 and references cited therein.  (2)  Compound 25 was fírst observed and characterized by E. Tran, and subsequently by Dr. Craig Pamplin. (Tran, E.; Pamplin, C. B.; Legzdins, P., unpublished observations.)  (3)  Tran, E.; Legzdins, P., unpublished observations.  (4)  Debad, J. D.; Legzdins, P.; Rettig, S. J.; Veltheer, J. E. Organometallics 1993, J2, 27142725.  (5)  a) Legzdins, P.; Rettig, S. J.; Sánchez, L. Organometallics 1988, 7, 2394-2403. b) Wada, K.; Pamplin, C. B.; Legzdins, P.; Patrick, B. O.; Tsyba, I.; Bau, R. J. Am. Chem. Soc. 2003,125, 7035-7048.  (6)  SIR92: Altomare, A . ; Cascarano, M . ; Giacovazzo, C ; Guagliardi, A . J. Appl Cryst. 1993, 26, 343.  (7)  Cromer, D. T.; Waber, J. T. International Tables for X-ray Crystallography; Kynoch Press: Birmingham, 1974; Vol. IV.  (8)  Ibers, J. A . ; Hamilton, W. C. Acta Crystallogr. 1964, 77, 781 -782.  (9)  Creagh, D. C ; McAuley, W. J. International Tables of X-ray Crystallography; Kluwer Academic Publishers: Boston, 1992; Vol. C.  (10)  Creagh, D. C ; Hubbell, J. H . International Tables for X-ray Crystallography; Kluwer Academic PubUshers: Boston, 1992; Vol. C.  (11)  CrystalClear: Versionl.3.5b20; Molecular Structure Corporation, 2002.  (12)  SHELXL97: Sheldrick, G. M . University of Gottingen, Germany, 1997.  Chapter 5. Mechanistic Insights into the Oligomerization of Cyclic Olefins by the Tungsten Nitrosyl Precatalysts^  A versión of this chapter has been submitted for pubHcation. Buschhaus, M . S. A.; PampUn, C. B.; Blackmore, I. J.; Legzdins, P. Transformations of Cyclic Olefms Mediated by Tungsten Nitrosyl Complexes. Reproduced in part with permission from Organometallics, submitted for publication. Unpublished work copyright 2008 American Chemical Society.  5.1 Introduction The four tungsten precatalysts illustrated in Scheme 5.1 produce ring-retaining oligomers of cyclohexene and other cyclic olefíns. The organic products are easily purifíed, but stable organometallic species are often harder to isolate, as has been discussed in Chapters 2 and 4 of this thesis. The nature of the catalytically-active tungsten species is of particular interest since under thermolysis conditions complexes 1 and 2 are known to form reactive alkylidene intermediates which might have been expected to react with the more strained cyclic olefins by olefín metathesis processes. This chapter rationalizes the observed reactivity of these tungsten complexes by proposing a mechanism for the catalytic, ring-retaining oligomerization of the cyclic olefins.  Scheme 5.1  In order to understand the nature of the catalytic species responsible for the cyclic olefin oligomerization, a number of diverse data and observations must be considered. These considerations begin with a detailed comparison of the reactivity of the tungsten complex 1 with its molybdenum analogue 3, and the insights into the alkylidene initiation pathway thus obtained. Then the initiation pathways of the other tungsten precatalysts are considered, and, based on comparison with 1, a convergent tungsten species is proposed. The oligomeric products are considered in light of this species, and a catalytic cycle is formulated that accommodates the known data. The explanatory power of the proposed mechanism is extended to other previously observed, but unexplained, cyclic-olefin couplings in Legzdins' systems. Finally, the possible decomposition routes of the purported catalytic species are considered.  5.2 Results and Discussion  5.2.1 Comparisons between the tungsten and the molybdenum systems Elucidation of the oUgomerization mechanism for the tungsten precatalysts begins with a consideration of the molybdenum system. As described in Chapter 3, Cp*Mo(NO)(CH2CMe3)2 (3) forms the reactive alkylidene intermediate at room temperature. The intermediate then reacts with the cyclic-olefin substrate to form a cis-metallacycle, which converts to a transmetallacycle, followed by an r|''-diene complex (see Scheme 3.1). The net resuh in terms of the organic ligand is a coupling of the neopentylidene ligand to the cyclic-olefin substrate along with a loss of dihydrogen. A comparison of the reactivity of the tungsten precatalyst 1 to that of the molybdenum complex 3 suggests strong initial similarities. Precatalyst 1 is known to form an alkylidene intermediate under thermolysis conditions of 70 °C.' This alkylidene reacts with the fivemembered rings of cyclopentene and 2,5-dihydrofuran via 2 + 2 addition to form the stable cismetallacycle complexes 25 and 29, respectively. These are analogous to the molybdenum cismetallacycle 7 formed from the reaction of 3 with cyclopentene (Scheme 5.2).  Scheme 5.2  25  29  7  In the presence of cyclohexene, 3 forms a metallacycle (17) en route to an t]''-diene complex (16). Similarly, the tungsten system of 1 gives ' H N M R evidence of a metallacyle (24) present in the complicated product mixture, although it can not be isolated (Scheme 5.3). The cyclohexene oligomers recovered from the thermolysis of precatalyst 1 in cyclohexene contain  small amounts of dimers and trimers capped with a neopentyl end-unit, suggesting an initiation step for 1 that parallels the formation of the coupled organic ligand in the molybdenum system.  Scheme 5.3  The reaction of 3 with cyclooctene proceeds slowly, needing heat to complete the transformation from the trans-metallacycle (9) to the r|''-diene complex (10). Similarly, reaction of 1 with cyclooctene gives organometallic products that are more stable than those obtained with cyclohexene (Scheme 5.4). Specifically, the structure of the trans-metallacycle 26 parallels that of the metallacycle of the molybdenum system. Products 27 and 28 show complete loss of the neopentyl ligands and the coupling of two cyclooctene molecules in the tungsten atom's coordinatíon sphere. Presumably, the tungsten center loses the coupled neopentyl-cyclooctene ligand from 26 and subsequently incorporales two cyclic olefin molecules to form 27 and 28.  3  1  For molybdenum, the final r[ -diene persists unless high temperature conditions are employed. Under these thermolysis conditions, the coupled organic can be released, and when cyclohexene is used as the solvent trace amounts of cyclohexene oligomers are observed. Intuitively, loss of the coupled organic ligand suggests that the [Cp*Mo(NO)] fragment remains; however, no molybdenum-containing products have been recovered. Additionally, the catalytic dimerization of allylbenzene necessitates the loss of a coupled organic containing the neopentyl group (vide infra, Scheme 5.9). In sum, the evidence suggests that the mechanistic pathway of 1 foUows that of the molybdenum analogue 3 but then readily proceeds ñjrther into the observed oligomerization reactivity. The initial reactive intermediate in both cases is an alkylidene complex. ' H N M R data confírm the formation of trans-metallacycles. This is followed by displacement of the coupled ligand, with the neopentyl-capped cyclohexene oligomers explaining the fate of the neopentyl fragment. Loss of the coupled ligand would leave the [Cp*W(NO)] fragment, presumably to  form the core of the active oHgomerization catalyst that incorporates the cyclic-olefin substrate. Therefore, this fragment will provide the basis for further considerations on the nature of the tungsten catalyst and its ability to oligomerize the cyclic-olefin substrates.  5.2.2 Altérnate oligomerization systems: Precatalysts 5 and 6 Other Legzdins' group compounds, specifically precatalysts 5 and 6, also oligomerize the cyclic olefins. Discussion here focuses on precatalyst 5. Precatalyst 6 is derived from 5 and gives the same reactive intermediate species (ri'-alkyne); therefore it has been studied less extensively. As described in Chapter 2, thermolysis of 5 in cyclohexene at 100 °C for 24 h affords a mixture of ring-retaining cyclohexene oligomers. In hopes of isolating a tungsten-containing species so as to gain some mechanistic insight, the reaction has been carried out at a lower temperature. Thermolysis of 5 in cyclohexene at 50 °C for 24 h produces a dark red oil, the ' H N M R spectrum of which shows the presence of oligomers and one major organometallic species. Column chromatography on silica foUowed by recrystallization fi-om Et20/hexanes affords 41 in 27% yield. The most probable mechanism for the formation of 41 is the coordination and coupling of cyclohexene with [Cp*W(NO)(ri'-HC=CPh)], the intermediate complex known to be formed by gentie thermolysis of 5.' Formation of the metallacycle would then be followed by the activation and transfer of a P-hydrogen to form 41 as depicted in Scheme 5.5. The solid-state molecular structure establishes that 41 is a 1,4-diene complex (Figure 5.1), and its characteristic spectroscopic properties confirm that it retains this molecular structure in solution.  Scheme 5.5  5  Figure 5.1 Solid-state molecular structure of 41 with 50% probability thermal ellipsoids shown. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.269(4), W(l)-C(2) = 2.249(3), W(l)-C(7) = 2.388(4), W(l)-C(8) = 2.303(4), C(8)-C(7) = 1.392(6), C(7)-C(6) = 1.516(5), C(6)-C(l) = 1.517(5), C(l)-C(2) = 1.396(9), W(l)-N(l) = 1.787(3), N(l)-0(1) = 1.229(4), W(l)-N(l)-0(1) = 175.4(3), C(8)-C(7)-C(6) = 123.3(3), C(7)-C(6)-C(l) = 100.1(3), C(6)-C(l)-C(2)= 120.8(4).  To confírm that 41 is indeed a precursor to the catalytic cycle and not merely a byproduct, an isolated sample has been dissolved in cyclohexene and heated overnight at 50 °C. Analysis of the fínal product mixture by ' H N M R spectroscopy reveáis the presence of cyclohexene oligomers and unreacted 41 (Figure 5.2), thereby indicating that it is indeed a precursor to the catalytically active species. Complex 41 can initiate oligomerization by dissociation of one of the olefínic linkages to open a coordination site at the metal center for the binding and coupling of cyclohexene or 1,4-cyclohexadiene. Evidence for this rationale is provided by the E I - M S data of the 1,4-cyclohexadiene oligomer mixture which exhibits peaks  attributable to oligomers containing a vinyl end group (CH=CHPh) analogous to the 1,4-diene extant in 41 (Chapter 2.2.2).  «IB  «  6  4  s  5  Figure 5.2 ' H N M R spectrum (300 MHz, CeDe, rt) of the reaction mixture of 41 in neat cyclohexene after thermolysis at 50 °C for 16 h, demonstrating the formation of cyclohexene oligomers and the remaining presence of unreacted 41. The arrows indícate diagnostic peaks due to 41, while the cyclohexene oligomers give rise to peaks between 1 - 2 ppm and 5.3-5.8 ppm.  As outlined in Chapter 2, the cyclohexene oligomers produced by precatalysts 5 and 6 are identical in form and distribution to those produced by precatalyst 1, with the exception of the small percentage of oligomers with capping end-groups. These end groups (ex. CH=CHPh or neopentyl) vary with the precatalyst employed and represent the initiation steps of the catalytic mechanism. This commonality in the products from different starting complexes suggests that variable initiation routes lead to a common catalytic species that is responsible for the cyclic olefín oligomerization. Examination of the precatalyst compounds shows that they all share the [Cp*W(NO)] core, with a variability in the organic ligands. Again, a catalytic species based on the [Cp*W(NO)] fragment is implied.  5.2.3 The proposed Cp*W(NO)(cyclic olefin)2 reactive species On the basis of comparisons between the tungsten precatalysts, their respective initiation reactions, and the molybdenum system, the [Cp*W(NO)] fragment is proposed as the basis for the catalytically active species. However, it is highly unlikely that this 14 electrón fragment would exist by itself in solution. Therefore, the solvated Cp* W(NO)(cyclic olefín)2 complex, as shown in Scheme 5.6, is proposed as a reasonable representation of the [Cp*W(NO)]-containing species. The coordinated olefins would then couple in the metal's coordination sphere to form a metallacyclopentane, a structural motif for which literature precedents exist.-'  Scheme 5.6  1  The coupled organic ligand of the tungstenacycle could undergo P-hydrogen eliminaüon to form a coordinated, mono-unsaturated cyclohexene dimer. On the simple tungstenacycle illustrated in Scheme 5.6 four p-hydrogens are available to the metal center, and due to the symmetry of the tungstenacycle, two possible dimer products could result. The newly opened space on the metal center would be filled by a third molecule of the cyclic-olefm substrate. The coordinated dimer can then be released, replaced at the metal center by a fourth olefin molecule to regenérate the Cp*W(NO)(cyclic olefin)2 complex. Altematively, the dimer can couple with the cyclic olefin to form a new tungstenacycle, and undergo P-hydrogen elimination to form a cyclohexene trimer. In an unsymmetrical metallacycle, such as occurs in the formation of trimers, up to four unique products might form at the elimination step. Scheme 5.7 summarizes the possible dimer and trimer products predicted to form from cyclohexene according to this proposed mechanism.  The experimentally observed cyclohexene oligomers correlate well with the general trends implicit in Scheme 5.7. As portrayed in the upper right of the scheme, two cyclohexene dimers are predicted, with the double bond in variable positions. The first of these predicted dimers matches the experimentally identified 3-cyclohexylcyclohexene (Chapter 2.2.7) which is the major mono-unsaturated dimer in the oligomer mixture. Up to twelve unique trimers are predicted by the proposed mechanism, which is strongly reminiscent of the family cluster made up of many trimer peaks observed in G C / M S analyses of the oligomers (Chapter 2, Figure 2.5). The ñirther complexity of the isolated cyclohexene oligomers (múltiple unsaturations) is accounted for by the previously established transfer dehydrogenation process (Chapter 2.2.3). The cyclohexene dimers and trimers capped with a neopentyl group result when the coupled neopentylcyclohexene from the initiation of precatalyst 1 remains coordinated to the metal center and subsequently couples to fiarther molecules of the cyclic-olefin substrate. When a similar process occurs with 4-methylcyclohexene as the cyclic substrate, peaks due to four unique neopentyl-capped dimers are evident in the GC/MS spectrum. The proposed mechanism suggests that these four compounds can be explained as the possible combinations for the position of the methyl groups relative to the neopentyl group. Thus, in the initial formation of the metallacycle that couples the neopentyl (in the form of the alkylidene) to the first 4methylcyclohexene molecule, the methyl group must be in one of two possible positions relative to the neopentyl, resulting in two products. In the subsequent coupling to the second molecule of 4-methylcyclohexene, the second methyl group also has two possible positions. In total, four combinations exist, leading to four peaks in the GC trace.  5.2.4 The proposed catalytic cycle for oligomerization of the cyclic olefíns Scheme 5.8 illustrates the proposed catalytic cycle for the oligomerization of cyclic olefins, beginning with the initiation pathways for precatalysts 1 and 5. The precatalysts form their respective reactive intermediates, couple an initial molecule of the cyclohexene substrate and lose the resultant organic ligand to form the common Cp*W(NO)(cyclic olefin)2 species. This species couples the two coordinated cyclohexene molecules in the metal's coordination sphere to form a tungstenacycle, which then undergoes p-hydrogen elimination to genérate a coordinated olefin. Finally, the coupled olefin may be replaced by cyclohexene to regenérate the  C p . W(NO)(cycl,c olefin), species. or al.ema,ive.y teher couplings may occur .o genérate ir.gher-„rder oligomers of cyclohexene, illustrated here with one specific trimer example. Scheme 5.8  o  0^7  O o 24 Neopentyl coupled to cyclohexene  41 CH=CHPh coupled to cyclohexene  5.2.5 The catalytic cycle applied to the dimerization of allylbenzene by 3 The observed reactivity of 3 with the allylbenzene substrate to form 23 can be explained by a mechanistic pathway analogous to that just described for the cyclic olefíns with precatalysts 1 and 5. Scheme 5.9 outlines the main steps, beginning with formation of 21 by reaction of one equivalent of allylbenzene with 3. The coupled diene ligand is released as 22, and two equivalents of allylbenzene coordínate to form a bis-olefin complex in which the steric interactions are minimized. Coupling of the olefíns in the metal's coordination sphere produces the fíve-membered metallacycle. Then P-hydrogen activation and reductive elimination produce the observed dimer 23, and the active species is regenerated by coordination of fiírther allylbenzene. The specifíc p-hydrogen activation illustrated is likely favored by the benzylic nature of that hydrogen. This, together with the minimization of steric interactions in the bisolefín, accounts for the formation of a single, specifíc allylbenzene dimer.'*  Scheme 5.9  3  5.2.6 Other examples of cyclic olefins coupled in the coordination sphere of Cp*W(NO) complexes The proposed mechanism for the coupling of cyclic olefms by a tungsten bis-olefm species outlined above in section 5.2.4 can be used to rationalize the observed reactivity of several other Legzdins systems for which mechanistic insight had been lacking. The fírst system involves the treatment of 1 with dihydrogen at rt, which results in the loss of neopentane to yield a putative Cp*W(NO)(CH2CMe3)(H) complex. When this reaction is carried out in cyclohexene as the substrate, three products are isolated, as illustrated in Scheme 5.10.-  Scheme 5.10  Product 42 has been previously synthesized by an altérnate route and is now identifíed by comparison of its characteristic spectral data.^ Products A l and 43 have been isolated and fuUy characterized, including X-ray crystallographic analyses. The solid-state structure of A l can be found in the Appendix of this thesis. The solid-state molecular structure of 43 (Figure 5.3) shows that two cyclohexene molecules have been coupled in the metal's coordination sphere, and it clearly displays the r|3,r|' binding motif of the resulting organic ligand. Compound 43 is proposed to form via a Cp*W(NO)(cyclohexene) intermediate (B), but no further mechanistic insight is offered in the original communication. A n isolated sample of 42 does not convert to 43  under the experimental conditions employed,^ so it cannot be invoked as an intermediate in the reaction mechanism. Also, use of D2 instead of H2 resuhs in no incorporation of deuterium in 42 and 43.  Figure 5.3 Solid-state molecular structure of 43 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.273(2), W(l)-C(8) = 2.284(2), W(l)-C(9) = 2.268(2), W(l)-C(10) = 2.434(2), W(l)-N(l) = 1.780(2), N(l)-0(1) = 1.225(3), C(l)-C(6) = 1.555(3), C(6)-C(7) = 1.559(3), C(7)-C(8) = 1.531(3), C(8)-C(9) = 1.413(3), C(9)-C(10) = 1.401(3), W(l)-C(l)-C(6) = 102.48(15), C(l)-C(6)-C(7) = 109.21(19), C(6)-C(7)-C(8)= 111.07(19), C(7)-C(8)-C(9) = 120.8(2), C(8)-C(9)-C(10) = 118.8(2), W(l)N(l)-0(1)= 174.81(18).  The formation of 43 from B can now be rationalized further, as illustrated in Scheme 5.11, by using the key intermediates of the proposed thermolysis-induced oligomerization mechanism. Intermedíate B readily adds a second equivalent of cyclohexene to give the bisolefin adduct. After initial coupling to form the tungstenacycle, a double P-hydrogen activation results in the loss of H i and the formaüon of the observed ri^,ri'-allyl alkyl ligand.  Scheme 5.11  A single product (44), similar in structure to 43, is obtained from the hydrogenolysis of 1 in 1,3-cyclohexadiene at rt.^ A n X-ray crystallographic analysis reveáis a ri^,ri' binding motif in the organic ligand analogous to that found in 43, but there is an additional uncoordinated double bond in the ligand, as illustrated in Figure 5.4.  Figure 5.4 Solid-state molecular structure of 44 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.272(2), W(l)-C(8) = 2.295(2), W(l)-C(9) = 2.283(2), W(l)-C(10) = 2.424(2), W(l)-N(l) = 1.7817(17), N(l)-0(1) = 1.225(2), C(l)-C(2)= 1.502(3), C(2)-C(3) = 1.334(3), C(l)-C(6) = 1.558(3), C(6)-C(7) = 1.566(3), C(7)-C(8) = 1.531(3), C(8)-C(9) = 1.417(3), C(9)-C(10) = 1.399(3), W(l)-C(l)-C(6) = 103.66(13), C(l)-C(2)-C(3) = 125.0(2), C(l)-C(6)-C(7) = 109.11(16), C(6)-C(7)-C(8) = 109.86(18), C(7)-C(8)-C(9) = 120.5(2), C(8)-C(9)-C(10) = 118.9(2), W(l)-N(l)-0(1) = 173.66(17).  Scheme 5.12 illustrates the proposed mechanism by which 44 forms. Again, a bis-cyclic olefin complex which then undergoes coupling to form a tungstenacycle accounts for the observed product. In order to form 44 no loss of H2 is necessary since a slight shift coordinates a double bond beside one of the alkyl links to form an t)'-allyl and thus satisfies the electronic needs of the metal center.  Scheme 5.12  A comparison of the characteristic H N M R signáis for the three cyclohexene-derived products formed from 1 under hydrogenolysis conditions with the ' H N M R spectra observed for the product mixtures obtained from 1 under thermolysis conditions reveáis a startling result. None of the three hydrogenolysis products are present in the thermolysis mixture. Product A l would not be expected since it is not accessible via an alkylidene intermediate. But products 42 and 43, proposed to form via B, should be accessible under thermolysis conditions as well. Considering the case of 43 first, it seems that the temperature difference induces a change in reactivity after the formation of the tungstenacycle. As illustrated in Scheme 5.13, in the room temperature system P-hydrogen activation occurs twice, resulting in H i loss and formation of tiie stable allyl-alkyl product. Conversely, at 70 °C the initially activated P-hydrogen is transferred from the metal back to the ligand to form a coordinated monounsaturated dimer, and the complex continúes oligomerization. If, however, under thermolysis conditions a second Phydrogen activation beside the alkyl linkage is envisioned, dimers with two unsaturations could be formed with loss of dihydrogen. The resulting complexes could also retum to the oligomerization cycle with loss of the dimer. Recall that the molybdenum trans-metallacycles such as 9 lose H2 to form the t)"'-diene complexes.  í  1 in cyclohexene under H2 atm, rt  1 in cyclohexene, thermolysis at 70 °C  -2H  or  dimer, one unsaturation; reform [W](cyclohexene)2 dimer, two unsaturations; reform [W](cyclohexene)2  [W] = Cp*W(NO)  Considering 42, comparisons of several systems again suggest that temperature plays a key role. As seen above in Scheme 5.10, 42 forms at room temperature when 1 undergoes hydrogenolysis in cyclohexene, presumably via intermediate B (Scheme 5.14, fírst equation). Compound 42 was originally synthesized by the reaction of Cp*W(NO)(r|^CH2CHCMe2)(CH2CMe3) with cyclohexane at 50 °C (second equation).^ A mechanism involving múltiple C - H bond activations of the cyclohexane substrate, coupling of the allylderived fragment with the cyclohexane-fragment, and loss of that coupled ligand via C-H bond activation of a second equivalent of cyclohexane is proposed in the publication. The fínal intermediate is proposed to be the cyclohexene adduct (B) which undergoes a C - H bond activation to form 42.' Finally, thermolysis reactions of 1 in cyclohexane at 70 °C show that C-H activations again cause the formation of B, which is then trapped in the presence of PMes (third  equation). However, in the absence of a trapping agent, decomposition is observed at 70 °C instead of the isolation of 42 (fourth equation). Thus, it is clear that 42 is thermally stable to 50 °C for 6 h, but that the higher temperatures employed in the oligomerization reactions (70 °C for 40 h) lead to the decomposition of any 42 formed.  Scheme 5.14  A second Legzdins system has been reported to couple two molecules of a cyclic-olefin substrate under hydrogenolysis conditions. In this case, a general method was developed to react Cp*M(NO)(CH2SiMe3)2 ( M = M o or W) with H2 and acyclic dienes at low temperatures to lose two equivalents of tetramethylsilane and genérate r|''-trans-diene complexes. However, when  1,3-cyclooctadiene was employed as the diene, the resulting product (45) had two molecules of the substrate coupled in the metal's coordination sphere. The organic ligand was determined to be 2-cyclooct-2-en-l-yl-1,3-cyclooctadiene, a triene coordinated to the metal center in a bis-r)' fashion, by single-crystal X-ray analysis. No mechanistic insight was possible at the time that these results were communicated.^ Now, application of the ideas of an initial bis-olefin complex followed by olefin-coupling in the metal's coordination sphere to form a metallacycle, p-hydrogen activation and rearrangement can rationalize this reactivity. Scheme 5.15 illustrates two possible routes by which the observed triene ligand may be formed.  Scheme 5.15  Finally, the mechanistic proposals that explain the cyclic-olefm oligomerization can rationalize the formation of 26, 27 and 28 during the thermolysis of 1 in cyclooctene, as well as the formation of cyclooctene dimers with one and two unsaturations (Chapter 4.2.3). Scheme 5.16 presents the proposed route to formation of 27 and 28. After the formation of the Cp*W(NO)(cyclic olefin)2 complex and subsequent conversión to the metallacycle, a double CH bond activation produces the 1,4-diene observed experimentally. Complex 27 in tum converts to 28, as shown by independent ' H N M R spectroscopy experiments (Chapter 4.2.3.2). Altematively, displacement of a cyclooctene dimer with either one or two unsaturations (double bonds) by the cyclooctene substrate reforms of the reactive Cp*W(NO)(cyclic olefm)2 complex.  Scheme 5.16  cyclooctene dimer, one unsaturation; and Cp*W(NO)(cyclic oIefin)2  cyclooctene dimer, two unsaturations; and Cp*W(NO)(cyclic olefin).  Compounds 27 and 28 are relatively stable with respect to temperature, and thus they can be isolated, whereas similar types of products are not observed or isolated when cyclohexene is the cyclic-olefm substrate. They can interconvert under thermolysis conditions, both in cyclic  olefín substrates and in CeDe. Finally, an isolated sample of 28 initiates ñirther oligomerization of cyclic olefíns under thermolysis conditions, demonstrating its ability to reenter the catalytic pathway. This likely occurs by loss of the diene ligand and coordination of two equivalents of the cyclic-olefm substrate to form the Cp*W(NO)(cyclic olefm)2 species, a process flxlly in keeping with the general mechanism proposed in this chapter. As seen in the later steps of Scheme 5.13 and Scheme 5.16, the formation of a cyclicolefm dimer with two unsaturations requires the loss of two equivalents of hydrogen. At present insufficient data exists to state defmitively in what form these hydrogens are lost. It may be as H2, as demonstrated by ' H N M R spectroscopy for the transformation from the trans-metallacycle to the ri'*-diene in the molybdenum system. Altematively, a direct metal-mediated transfer to the cyclic-olefin substrate may be occurring, as suggested by the ' H N M R data described in Chapter 2.2.3 for the cyclohexene and 1,4-cyclohexadiene oligomerizations with precatalyst 5. On the other hand, the transfer dehydrogenation process described in Chapter 2 may be facilitated by the catalytically active species in reaction steps that are independent of the steps of the oligomerization. At this point, no conclusions on the nature of the " 2 H " lost are supported by the available experimental data.  5.2.7 Comparison to historical cyclohexene oligomerization reactions The proposed catalytic cycle is based on a Cp*W(NO)(cyclic olefm)2 species which subsequentiy forms a tungstenacycle. In contrast, the 1970-80s oligomerizations of cyclohexene using tungsten-based metathesis catalysts are limited in their mechanistic insight.^ Only one report suggests a mechanism, said to involve a tungsten hydride species with which a cyclohexene molecule forms an adduct and subsequentiy reacts.^'' In contrast, the mechanisms proposed for the Legzdins systems described here do not contain any such species with both a coordinated cyclohexene molecule and a hydride. The transient hydride species proposed to form in the coupling of the cyclic olefins contain either an alkyl or an allyl linkage to the organic ligand. The stable hydride-containing compoimds which have been isolated (8, 28) are all allyl hydrides. Thus, hydride species analogous to those proposed in 1980 are neither observed ñor invoked in the present system.  5.2.8 Mechanistic proposals for the reaction of 1 and 5 with the oxygen-containing cyclic olefíns Oxygen-containing cycUc-olefín substrates tend to undergo ring-opening rather than oligomerization reactions with the tungsten precatalysts. Mechanistic explanations proposed here employ structures similar to those invoked for 1 and 5 with the cyclic olefms, particularly in the preliminary formation of a metallacycle. Complex 1 reacts with the five-membered ring of 2,5-dihydrofuran to form a cismetallacycle (29). This is the expected product since in both the tungsten and molybdenum systems the small ring size forms a stable product and prevents ftirther reactivity. In contrast, 1 reacts with 3,4-dihydro-2H-pyran to form an alkoxy allyl complex in which the ring has been opened (30). As Scheme 5.17 proposes, the coupling of the alkylidene intermediate o f l with the olefín forms the metallacycle. A P-hydrogen activation and transfer analogous to the cyclohexene system forms a coordinated olefín that also contains a metal-oxygen interaction. Cleavage of the C - 0 bond forms the alkoxy allyl complex observed.  Scheme 5.17  30  In a similar maimer, the reactivity of 5 with 3,4-dihydro-2H-pyran can be rationalized by coupling of the reactive ri^-alkyne intermediate with the substrate to form a metallacycle.  Subsequent (3-hydrogen activation and transfer forms a coordinated olefin whose heteroatom can interact with the metal center. The alkoxy allyl product 31 forms through C - 0 bond cleavage as shown in Scheme 5.18.  Scheme 5.18  5  31 In contrast, the reaction of 5 with 2,5-dihydrofuran seems anomalous due to the atypical configuratíon observed in the final product 32. However, a speculative mechanism can be proposed. The expected initial step would be formation of a metallacycle through coupling of the intermediate species with the substrate. It is possible in this complex, as drawn in Scheme 5.19, for the oxygen atom to interact with the metal center, and thus become susceptible to a C-O bond cleavage and rearrangement that produces the observed product. A similarly-opened product is formed in an isoelectronic Cp2Zr(3-methoxybenzene) system, where the proposed mechanism is metallacycle formation by coupling of 2,5-dihydrofuran with the 3-methoxybenzene, followed by C - 0 bond cleavage."' The difference in reactivity with precatalyst 1 is presumably due to the inability of the oxygen atom to approach the metal center because of the rigidity of the cismetallacyle.  5  \  32  5.2.9 Decomposition of the oligomerization catalyst: Concentration effects revisited As presented in Chapter 2, Table 2.4, the yield of cyclohexene oligomer produced by the tungsten precatalysts varies with the initial solution concentration. Decreasing the initial concentration of 1 causes an increase in the number of moles of cyclohexene converted per mole of precatalyst, with concentrations of 0.005 M - 0.010 M giving optimal conversions. Figure 5.5 illustrates the general trend for representative samples of 1 thermolyzed in cyclohexene for 40 h at 70 °C. The samples demónstrate the trend to higher substrate tumover at lower concentrations.  100  o  0.01  0.02  0.03  0.04  0.05  Concentration (M)  Figure 5.5 The effect of variable initial concentrations of 1 in cyclohexene on catalyst activity (tumover number, añer 70 °C, 40 h). The trend Une is drawn to guide the eye only, and error bars are calculated based on imcertainties of ± 2 mg of 1 and ± 1 0 mg of oligomer recovered.  The inverse trend observed suggests that lower concentrations discourage a catalyst decomposition pathway. Any oligomer-forming reaction would be expected to show a positive, proportional relationship between precatalyst levéis and total oligomer product formation, and the ratio of moles of oligomer formed per mole of precatalyst should not change in a given unit of time. The observed trend shows that another reaction (or perhaps several) is occurring that diverts the catalyst out of the catalytic cycle. If this catalyst decomposition occurred intramolecularly, the probability of its occurrence would be related to the relative rates of reaction, not the concentration of the rest of the solution. If the decomposition is intermolecular, with two tungsten-containing species coming together, then concentration dependence would be expected since increasing the concentration of the tungsten species would increase the probability of two such species meeting. The result would be a decrease in oligomer formation as the concentration of tungsten precatalyst increases, as is observed. The decomposition product(s) of the Cp* tungsten precatalysts have not been identified. However, for the Cp tungsten system, several decomposition products of 2 have been formulated as bimetallic species (Chapter 4.2.9.1). Compounds 36, 37, 38 and 39 are not likely candidates for a direct comparison to the Cp* system; however, they demónstrate the feasibility of a dimeric tungsten species. Similarly, compound A l in Scheme 5.10 shows the feasibility of a bimetallic tungsten species with Cp* ligands. A potential decomposition product in the oligomerization reactivity of 1 may be (Cp*W(N0))2.  The tungsten and molybdenum catalysts of Schrock also exhibit bimolecular decomposition pathways. Two modes of decomposition are observed in Schrock systems when ethylene is the metathesis substrate, although neither reaction has been elucidated in detall. First, a bimolecular coupling of the reactive alkylidene complex, particularly for methylene complexes, forms an ethylene complex that is no longer reactive to metathesis. Second, rearrangement of the metallocyclobutane intermediate leads to olefín formation by p-hydride elimination, and then to an ethylene complex identical to that formed by the fírst process. In the presence of substrates such as 2-pentene, bimetallic decomposition products are formed, those of the tungsten system having unsupported W=W bonds and those of the molybdenum system having bridging imido groups." In summary, both the experimental evidence and comparisons to related systems in the literature support the plausibility of a bimetallic decomposition pathway for the reactive  Cp*W(NO)(cyclic olefín)2 species and its derivatives. The number of operative pathways can not be determined, ñor can the details of the decomposition reaction(s) be elucidated. The nature of the final bimetallic species is unknown, although the (Cp*W(N0))2 complex is suggested.  5.3 Summary A mechanistic rationale for the catalytic oligomerization of cyclic olefms by thermolysis of precatalysts 1, 2, 5 and 6 in the olefin substrates has been presented. The initiation of precatalysts 1 and 5, and also of 3 in the dimerization of allylbenzene, involves the coupling of one equivalent of the substrate with the reactive intermediate typically formed by the precatalyst. The coupled ligand rearranges to an olefin, or to a diene with loss of H2, and is subsequently released from the metal center. The metal then coordinates two equivalents of substrate to form a bis-olefin complex. The bis-olefin formed from precatalysts 1 and 5, Cp*W(NO)(cyclic olefin)2, represents the convergent entry point to the catalytic cycle for both precatalysts. The coordinated olefins undergo metal-mediated coupling to form a metallacyclopentane complex. The metallacycle then undergoes P-hydrogen elimination and reductive elimination to genérate a cyclic-olefin dimer as a coordinated olefin. Further addition and coupling of substrate leads to formation of trimers and higher oligomers. Altematively, loss of the coordinated oligomer regenerates the reactive bis-olefm complex. The feasibility of this mechanism is supported by the available cyclic-olefin data from both the molybdenum and timgsten systems. The mechanism is further supported by its ability to rationalize a range of couplings observed in Legzdins systems. The dimerization of allylbenzene by 3 yields a specific product (23) that is consistent with the proposed mechanism. The previously observed coupling of cyclohexene, 1,3-cyclohexadiene and 1,3-cyclooctadiene under hydrogenolysis conditions at ambient temperatures can now be explained as occurring via variations of the thermolysis mechanism. The isolated products from the thermolysis of 1 in cyclooctene also fit within the explanatory power of the proposed mechanism. Finally, decomposition for the tungsten catalyst species is consistent with a bimetallic pathway, the details of which have not been determined. The effect of this bimetallic decomposition is observed in the increase of cyclic-olefm conversión with decreasing precatalyst concentration.  5.4 Experimental Procedures  5.4.1 General Methods For applicable general methods, consuk section 4.4.1 in Chapter 4 of this thesis. Complex 41 was made by Dr. lan Blackmore. Details of its synthesis and characterization data are included here for completeness.  5.4.2 Preparation of 41 A re-sealable reaction vessel was charged with Cp*W(NO)(CH2SiMe3)(ri'-CPhCH2) (5) (350 mg, 0.65 mmol) and cyclohexene (8 mL). The red solution was then heated at 50 °C for 16 h, and the volatile components were removed from the final reaction mixture under reduced pressure to obtain an oily red solid. This solid was redissolved in a mínimum of pentane, and the solution was transferred to the top of a column of silica (ca. 6 x 0.5 cm). The organic products were eluted with a large volume of pentane, and the remaining red-brown band was eluted with 3:1 pentane/Et20 and the eluate was collected. Removal of the volatile components from the eluate and re-crystallization of the residual solid from a mixture of Et20/hexanes at -30 °C afforded complex 41 as bright red crystals (93 mg, 0.17 mmol, 27%). Anal Caled for C24H31NOW (533.35): C, 54.05; H , 5.86; N , 2.63. Found: C, 54.14; H, 5.93; N , 2.91. IR (Nujol): VNO 1588 cm''. M S (EL 120 °C): m/z 535. ' H N M R (CeDe) 5; 7.20-7.18 (m, 3H, aryl), 6.98-6.92 (m, I H , aryl), 4.61 (m, IH, PhC//=CH), 3.14, 2.65, 1.72 and 1.96 (m, I H each, (overiapping) cyclohexenyl-C7^2), 2.22 (m, IH, olefmic CH), 1.83 (dd, I H , (overiapping) olefinic CH), 1.59 (m, 2H, (overiapping) cyclohexenyl-C//2), 1.41 (s, 1 5 H , CjMes), -0.97 (t, I H , olefinic-C//), cyclohexenyl-allylic protón obscured. '^C N M R (CeDg) 5;  146.7  (ipso-aryl),  128.6, 126.3, 124.7  (ortho, meta and para-ary/),  102.1 (CsMes), 69.2, 48.5, 45.2, 11.2 (olefinic carbons), 30.6 (cyclohexenyl-allylic carbón) 28.4, 27.7 17.7 (cyclohexenyl-CH2), 9.3 (CjA/es).  5.4.3 ' H N M R spectrum of 41 in cyclohexene after heating at 50 ° C for 16 hours A sample of 41 ( 1 0 mg) was dissolved in cyclohexene and heated at 5 0 °C for 16 h. The volatile components were removed in vacuo and the residue analyzed by ' H N M R spectroscopy.  5.4.4 X-ray Crystallography. Data coUection was carried out at -100 ±1 °C on a Bruker X8 A P E X diffractometer, using graphite-monochromated Mo K a radiation. Data for 41 were coUected to a máximum 19 valué of 55.8 ° in 0.5 ° osciUations. The structure was solved by direct methods'^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l , H2, H7 and H8 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of ftiU-matrix least-squares analysis was based on 3957 observed reflections and 265 variable parameters. Data for 43 were coUected to a máximum 19 valué of 57.0 ° in 0.5 ° osciUations. The structure was solved by direct methods'^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of fiaU-matrix least-squares analysis was based on 4813 observed reflections and 231 variable parameters. Data for 44 were coUected to a máximum 19 valué of 55.8 ° in 0.5 ° osciUations. The structure was solved by direct methods'^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H2, H3, H8, H9 and HIO were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 4467 observed reflections and 251 variable parameters. Neutral-atom scattering factors were taken from Cromer and Waber.Anomalous dispersión effects were included in Fcaic;'"* the valúes for A / and A / ' were those of Creagh and McAuley.'^ The valúes for mass attenuation coefficients are those of Creagh and Hubbell.'^ A l l calculations were performed using S H E L X L - 9 7 . ' X - r a y crystallographic data for the three structures are presented in Table 5.1, and in the cif files.  Table 5.1 X-ray Crystallographic Data for Complexes 41, 43 and 44. 41  43  44  Empirical fornmla  C24H31NOW  C22H33NOW  C22H31NOW  Crystal Habit, color  Prism, red  Prism, red  Needle, pink  Crystal size (mm)  0.60 x 0.25 x 0.10  0.2x0.2x0.1  0.40 x 0.07 X 0.04  Crystal system  Orthorhombic  Monoclinic  Triclinic  Space group  Pna2]  C2/C  P-i  Volume (Á^)  2049.84(15)  3868.93(10)  948.71(14)  a (A)  16.8210(7)  22.1618(3)  8.1917(6)  biA)  12.6397(5)  13.0316(2)  9.7406(9)  ciA)  9.6412(4)  17.9373(3)  13.6903(13)  aC)  90  90  73.624(3)  J3C)  90  131.682(2)  89.272(3)  ril z  90  90  65.724(3)  4  8  2  Density (calculated) (Mg/m^)  1.728  1.756  1.783  Absorption coefficient (mm'')  5.649  5.981  6.098  Fooo  1056  2032  504  Measured Reflections: Total  14882  66749  26325  Measiu-ed Reflections: Unique  3957  4813  4467  Final R índices'^  R l = 0.0178, wR2  R l = 0.0168, wR2  R l = 0.0152, wR2  = 0.0386  = 0.0379  = 0.0335  1.035  1.068  1.065  1.571 and -0.735  1.829 and -0.634  1.184 and -0.574  Crystal Data  Data CoUection and Refinement  Goodness-of-fit on  *  Largest diff. peak and hole (e'  R l on F = E I (|Fo| - \FS | / I |Fo|, (/> 2a(7)); wR2 = [ (E ( Fo' - F,' )^) / E w(F,' )'f' data); w = [ cr^Fo^ ]''; * GOF = [ E (w ( |Fo| - |F,| f ) / degrees offreedom ] ^'^.  (all  5.5 References (1)  Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, 223-233 and references therein.  (2)  Debad, J. D.; Legzdins, P.; Lumb, S. A.; Rettig, S. J.; Batchelor, R. J.; Einstein, F. W. B. Organometallics 1999,18, 3414 and references cited therein.  (3)  a) for proposals of intermediate metallacyclopentanes, see McLain, S. J.; Schrock, R. R. J. Am. Chem. Soc. 1978, i 1 3 1 5 - 1 3 1 7 and You, Y . ; Wilson, S. R.; Girolani, G . S. Organometallics 1994,13, 4655-4657. b) for examples of isolable metallacyclopentanes, see Ison, E. A.; Abboud, K . A.; Boncella, J. M . Organometallics 2006, 25, 1557-1564 and Amdt, S.; Schrock, R. R.; Müller, P. Organometallics 2007, 26, 1279-1290.  (4)  Graham, P. M . ; Buschhaus, M . S. A.; Pamplin, C. B.; Legzdins, P. Organometallics, 2008,27,2840-2851.  (5)  Jin, X . ; Legzdins, P.; Buschhaus, M . S. A. J. Am. Chem. Soc. 2005,127, 6928-6929.  (6)  Ng, S. H . K.; Adams, C. S.; Hayton, T. W.; Legzdins, P.; Patrick, B. O. J. Am. Chem. Soc. 2003,125, 15210-15223.  (7)  Jin, X . ; Legzdins, P., unpublished results.  (8)  Debad, J. D.; Legzdins, P.; Young, M . A.; Batchelor, R. J.; Einstein, F. W. B. J. Am. Chem. Soc. 1993,115, 2051-2052.  (9)  a) Moulijn, J. A.; van de Nouland, B. M . React. Kinet. Catal. Lett. 1975, 3, 405-408. b) Giezynski, R.; Korda, A . J. Mol. Catal 1980, 7, 349-354.  (10)  Cuny, G. D.; Buchwald, S. L . Organometallics 1991,10, 363-365.  (11)  Schrock, R. R.; Czekelius, C. Adv. Synth. Catal 2007, 349, 55-77 and references cited therein.  (12)  SIR92: Altomare, A . ; Cascarano, M . ; Giacovazzo, C ; Guagliardi, A. J. Appl Cryst. 1993, 26, 343.  (13)  Cromer, D. T.; Waber, J. T. International Tables for X-ray Crystallography; Kynoch Press: Birmingham, 1974; Vol. IV.  (14)  Ibers, J. A . ; Hamilton, W. C. Acta Crystallogr. 1964,17, 781-782.  (15)  Creagh, D. C ; McAuley, W. J. International Tables of X-ray Crystallography; Kluwer Academic Publishers: Boston, 1992; Vol. C.  (16)  Creagh, D. C ; Hubbell, J. H. International Tables for X-ray Crystallography; Kluwer Academic Publishers: Boston, 1992; Vol. C.  (17)  SHELXL97: Sheldrick, G. M . University of Gottingen, Germany, 1997.  Chapter 6. Thesis Summary and Future Dírections  6.1 Thesis summary This thesis presents the investigations carried out to elucídate the nature, extent, and mechanism of the cyclic-olefm oligomerization reactivity observed with a series of tungsten precatalysts, with particular focus on complexes Cp*W(NO)(CH2CMe3)2 (1) and Cp*W(NO)(CH2SiMe3)(Ti'-CPhCH2) (5). Under thermolysis conditions these tungsten precatalysts oligomerize simple cyclic olefíns, from cyclopentene to cyclooctene, into ringretaining oligomers as high as dodecamers (depending on the substrate) with remaining sites of unsaturation. Chapter 2 describes in detall the oligomeric products obtained when the cyclicolefín substrates are cyclohexene and 1,4-cyclohexadiene. Chapter 3 presents the reaction pathway of Cp*Mo(NO)(CH2CMe3)2 (3) with cyclic olefíns in order to compare it to the reactivity of 1, the tungsten analogue of 3. Chapter 4 explores the reactivity of precatalyst 1 with a range of cyclic olefíns, with an emphasis on the detectable and isolable organometallic complexes that give insight into the oligomerizaion mechanism. Finally, Chapter 5 culminates in the proposal of a catalytic cycle that provides a mechanistic explanation for the oligomerization reactivity described in Chapters 2-4 and also for a range of other cyclic-olefm coupling reactions previously observed in Legzdins group chemistry.  6.1.1 Precatalyst initiation The mechanism of the cyclic-olefm oligomerization reaction is composed of a distinct initiation pathway for each precatalyst, a catalytic cycle that produces the oligomers, and a decomposition reaction (or reactions) that destroys the active catalytic species. The initiation pathway of precatalyst 1 is elucidated through comparison to the related molybdenum complex, Cp*Mo(NO)(CH2CMe3)2 (3). Both compounds genérate reactive alkylidene intermediates of the form [Cp*M(N0)(=CHCMe3)] ( M = W or Mo) through hydrogen transfer and loss of neopentane. The alkylidene intermediates then react with a substrate cyclic-olefm molecule via a 2 + 2 addition to form a metallacyclobutane. In the molybdenum system, the cis-metallacycle derived from cyclopentene (7) can be isolated, and it subsequently forms a stable allyl hydride complex (8). The cis-metallacycles derived from larger ring sizes isomerize to trans-metallacycles (9,14,17), presumably via an allyl hydride intermediate analogous to 8. During the thermolysis of 1, analogous tungsten trans-metallacycles derived from cyclohexene and cyclooctene (24, 26) are detected by ' H N M R spectroscopy; however, the putative preceding cis-metallacycles have not been observed. The molybdenum  trans-metallacycles convert to ri'*-diene complexes with loss of dihydrogen (10,15,16), and, under thermolysis conditions, the coupled organic ligand will dissociate from the metal center and allow a small amount of cyclic-olefm oligomerization. In contrast, the tungsten transmetallacycles cannot be isolated since vmder the thermolysis conditions required to genérate the alkylidene the resulting metallacycle converts to a coordinated olefín by P-hydrogen transfer. Subsequent loss of the olefín and coordination of two equivalents of the cyclic substrate forms the proposed catalytic species Cp*W(NO)(cyclic olefin)2. Thus, as summarized in Scheme 5.1, the initiation of precatalyst 1 foUows a reaction pathway analogous to that outlined for molybdenum complex 3 but then readily proceeds into the catalytic oligomerization of cyclic olefms.  Scheme 6.1  1  n=l-4  Initiation of precatalyst 5 begins with formation of the reactive r)^-alkyne intermediate [Cp*W(NO)(ri^-HC=CPh)] by hydrogen transfer and loss of tetramethylsilane. The ri'-alkyne couples with a molecule of the cyclic olefín to form a metallacycle which then converts to a 1,4diene complex, an example of which has been isolated (41). Loss of the 1,4-diene ligand and  coordination of two equivalents of the cyclic substrate form the Cp*W(NO)(cyclic olefm)2 complex (Scheme 6.2).  Scheme 6.2  6.1.2 The catalytic cycle The proposed Cp*W(NO)(cyclic olefm)2 complex represents the entry point into the catalytic cycle, as illustrated in Scheme 6.3. In the tungsten's coordination sphere the cyclic olefíns couple to form a metallacyclopentane. Subsequent P-hydrogen activation and transfer yields a dihapto cyclic-olefin dimer, which may then either couple to a third molecule of the substrate or be displaced to regenérate the initial bis-olefin complex. hi this way, ring-retaining oligomers of various lengths and configurations are obtained. Consistent with the mechanism presented, the major cyclohexene dimer in the oligomer mixture is 3-cyclohexylcyclohexene. The detection of cyclohexene oligomers capped with neopentyl or CH=CPh end groups indicates that the coupled olefins formed during initiation of precatalysts 1 and 5 particípate in the coupling reactions in the catalytic cycle. While none of the species in the catalytic cycle have been directly observed, the coupling of the cyclic olefins in the coordination sphere of the tungsten complex is fiírther substantiated by the isolation of the 1,4-diene complex 27 and the allyl hydride complex 28, in which two molecules of cyclooctene have been coupled together. Tumover frequencies for precatalysts 1, 5 and 6 range from 5.5 to 6.5 mol/h at concentrations of 0.01 M in cyclohexene and 100 °C over a 24 h reaction time. Precatalyst 1 requires a mínimum temperature of 70 °C in order to readily form the initial alkylidene intermediate; the higher temperature of 100 °C increases both tumover and tumover frequency for the oligomerization of cyclohexene.  1,5  The coupling mechanism proposed to explain the oligomerization of simple cyclic olefms by precatalysts 1 and 5 can also be extended to rationalize the formation of other coupled products obtained under thermolysis or hydrogenolysis conditions. Precatalyst 1 under dihydrogen atmosphere at rt in the presence of cyclohexene or 1,3-cyclohexadiene forms allylalkyl complexes.' Under similar conditions Cp*M(NO)(CH2SiMe3)2 ( M = W or Mo) in 1,3cyclooctadiene forms an r)''-diene compound.^ In both cases, invoking the formation of a bisolefm complex that undergoes coupling to form a metallacyclopentane, analogous to the steps of the catalytic cycle above, rationalizes the formation of the observed products. Compound 3 catalytically dimerizes allylbenzene to yield a single product,3 and this selectivity is explained by the formation of a bis-olefm complex that minimizes the steric interactions of the coordinated allylbenzene substrate, followed by coupling and subsequent reléase of the organic product.  6.1.3 Catalyst decomposition and substrate limítations The active tungsten oligomerization catalyst is proposed to decompose via a bimetallic reaction pathway. The activity of precatalyst 1 in cyclohexene is concentration dependent, with greater substrate tumover achieved at lower precatalyst concentrations, which can be rationalized by such a decomposition mechanism. Heteroatoms within the cyclic-olefin ring, such as ether and amine iiinctionalities, are detrimental to the oligomerization pathway. Oxygen-containing 2,5-dihydrofuran and 3,4dihydro-2H-pyran react with 1 and 5 to form metallacycles in a manner analogous to the initiation pathways described in 6.1.1 (vide supra), followed by preferenfial acfivation of the accessible C - 0 bonds by the metal center to form ring-opened alkoxy products (30, 31, 32). The N - H bond of 1,2,3,6-tetrahydro-pyridine is activated by the alkylidene intermediate of 1 to form an amido product (33). In addition, the presence of a methyl group on the cyclic olefin leads to competing reactions. Oligomerization is thus currently limited to simple cyclic olefins.  6.1.4 Comparison of the Legzdins alkylidene complexes (Cp*M(NO)(=CHCMe3)) to known olefín metathesis alkylidene complexes The oligomeric products obtained from thermolysis of the tungsten precatalysts in cyclic olefins are ring-retaining. There is no evidence for R O M P processes in the systems examined in this thesis. As a result of ring-coupling, many isomers of the trimers and longer oligomeric lengths are formed, as evidenced by the number of peaks detected in the G C / M S analyses. In addition, the cyclohexene oligomers contain higher amounts of unsaturation than any related cyclohexene oligomer described to date in the literature. The high levéis of unsaturation observed are due to transfer dehydrogenation between the oligomer chains and the cyclic substrates, presimiably mediated by a tungsten species. Two early olefin metathesis experiments conducted in 1975 and 1980 report the ringretaining oligomerization of cyclohexene. These experiments, based on WCU in conjunction with a variety of addifives to form catalytically-active reacüon "soups", give oligomers similar to those obtained with precatalysts 1 and 5, but with much lower levéis of unsaturation.'* The reports provide very little mechanistic insight, but later work demonstrates that olefín metathesis is not the mechanism. In contrast, the oligomerization reactions described in this thesis begin with discrete precatalyst complexes whose initiation pathways can be determined, and the proposed catalytic cycle is supported by a variety of observations including several isolated  organometallic complexes containing two substrate molecules coupled in the metal's coordination sphere. The higher unsaturation levéis observed in the oligomers obtained from 1 and 5 introduce possibilities for functionalization at the olefinic sites. However, i f utilized in an industrial context, these products will occupy only a specialized niche market. It is of interest that 1, which forms a reactive alkylidene intermediate upon thermolysis, produces oligomers of cyclic olefins such as cyclopentene and cyclooctene but shows no signs of ROMP activity. The crucial involvement of alkylidene complexes in olefin metathesis has been known for many years and is amply demonstrated in the metathesis catalysts of Schrock and Grubbs.^''' Comparisons between the group-six Schrock-alkylidene complexes and those of the Legzdins group reveal several key differences that ultimately result in very different modes of reactivity. The Schrock alkylidenes can be isolated and fuUy characterized, and during metathesis reactions some of the alkylidene complexes propagating the polymerization can be observed by N M R spectroscopy. Metallacyclobutane complexes are proposed as intermediate species but they are generally not detectable.^ In contrast, the Legzdins alkylidenes derived from 1 and 3 under thermolysis conditions are the undetectable, highly reactive intermediates^ and the metallacycles resulting from coupling with cyclic olefins are the isolable or spectroscopically detectable species (vide supra). Both the Schrock and the Legzdins complexes are electrón defícient; however, the Schrock metathesis catalysts are formally d^, while the Legzdins precatalysts are formally á^. Thus, the Legzdins complexes are relatively electrón rich at the metal center, in large part because the nitrosyl ligand stabilizes the electrón rich d'' metal. These differences give the Legzdins alkylidene complexes their unique reactivity.  6.L5 Significance and impact The research presented in this thesis provides unique insights into two major áreas; (1) the mechanism for formation of ring-retaining oligomers from cyclic olefins, with particular focus on cyclohexene and the tungsten nitrosyl precatalysts 1 and 5, and (2) the reactivity of the alkylidene intermediates formed from complexes 1 and 3 with cyclic olefins in a manner distinct from the R O M P reactivity commonly observed in other tungsten and molybdenum alkylidene systems. In terms of the advancement of Legzdins' group chemistry, this thesis contributes an extensive understanding of a new pathway of reactivity for a long-established compound. The CH bond activation ability of 1 has been known for some time; now its reactivity with cyclic olefins has been described. In the process of doing so, the reactions of 2, 3, 4 and 5 with cyclic  olefins have also been explored. In addition, the isolation and characterization of oligomers expands the focus of the group to consider the importance of organic products. The mechanism proposed in this thesis provides a unified explanation for the observed precatalyst initiation reactions and the catalytic cyclic-olefin oligomerization.  6.2 Future dírections Several Unes of experimentation can be envisioned for ñiture exploration, including an expansión of the possible substrates, further refinement of mechanistic understanding, and reduction of catalyst decomposition. Currently the substrates for oligomerization compatible with precatalysts 1 and 5 are limited to simple cyclic olefins. The larger ring sizes form relatively stable organometallic products compared to those of cyclohexene, and thus it will be increasingly difficult to effectively oligomerize substrates beyond cyclooctene. Ether and amine groups within the substrate ring have been shown to be detrimental to oligomerization. Further experiments should test a wider range of fimctional groups for compatibility with the catalytic species. Ketones and polycyclic substrates might give interesting results. Substrates with readily activated C-H bonds will likely induce a competition between C-H bond activation and the desired oligomerization reactivity. Given the dimerization observed with 3, allylbenzene must be tested with the tungsten precatalysts despite the possibility of competitively activating the aryl hydrogen bonds. A selection of other acyclic olefins with available P-hydrogens could be tried as well. Mixed substrate systems could be explored, such as co-oligomerization of cyclohexene and cyclooctene. In terms of reaction condition optimization, the ideal concentration of precatalyst 1 in cyclohexene has been roughly determined. Further optimization should be done to refine the ideal reaction temperature and time conditions for both 1 and 5. Attempts to further probé the oligomerization mechanism in search of support (or lack thereof) for the proposed catalytic cycle will be difficult. The detection or the trapping of one of the species within the cycle would be ideal. More likely, the [Cp*W(NO)] fragment would be trapped, the trapping agent having displaced the coordinated cyclic olefins. The cholee of trapping agent will be of key importance since most conventional reagents such as phosphines will trap the alkylidene or ri'-alkyne intermediates as they form. One potential possibility may  be 1,3-butadiene, which should react with the [Cp*W(NO)] fragment to form the previously synthesized, stable rj^'-trans-diene complex Cp*W(NO)(Ti''-CH2=CHCH=CH2).^ The 1,3butadiene may, however, couple to the intermediates formed during initiation in a marmer analogous to the reaction of 3 with acyclic olefms (Chapter 3, section 3.2.15). Kinetic studies will be hampered by the complexity of the reaction mixture and the uncertainty about the exact nature of the catalyst decomposition products. The catalyst lifetime might be extended by attaching the complex to some solid or polymeric support, such that the active species would be prevented from forming the bimetallic species proposed for the decomposition pathway. A reasonable site of attachment would be a tether replacing one of the methyl groups on the Cp* ligand of the precatalyst. In fact, similar catalyst supports have been proposed to prevent bimolecular decomposition in metathesis systems.^'' A n additional benefit would be the easy separation of the catalyst matrix from the oligomeric products and remaining cyclic-olefm substrate. Of course, a change in the reaction environment of this magnitude will inevitably affect the oligomerization reactivity of interest, whether by increasing, decreasing, halting, or entirely changing the observed outcome.  6.3 References (1)  Jin, X . ; Legzdins, P.; Buschhaus, M . S. A. J. Am. Chem. Soc. 2005,127, 6928-6929, and unpubhshed observations (Jin, X . ; Legzdins, P.).  (2)  Debad, J. D.; Legzdins, P.; Young, M . A.; Batchelor, R. J.; Einstein, F. W. B. J. Am. Chem. Soc. 1993,115, 2051-2052.  (3)  Graham, P. M . ; Buschhaus, M . S. A.; Pamphn, C. B.; Legzdins, P. Organometallics, 2008,27,2840-2851.  (4)  a) Moulijn, J. A . ; van de Nouland, B. M . React. Kinet.Catal. Lett. 1975, 3, 405-408, and b) Giezynski, R.; Korda, A . J. Mol Catal 1980, 7, 349-354.  (5)  For general references see a) Olefin Metathesis and Metathesis Polymerization; Ivin, K . J.; Mol, J. C.; Academic Press, San Diego, 1997 and references cited therein, and b) Handbook of Metathesis; Grubbs, R. H., Ed.; Wiley-VCH, Weinheim, 2003, vols. 1-3: Catalyst Development (vol. 1); Applications in Organic Synthesis (vol. 2); Applications in Polymer Synthesis (vol. 3) and references cited therein.  (6)  a) Schrock, R. R.; Czekelius, C. Adv. Synth. Catal 2007, 349, 55-77 and references cited therein. b) Schrock, R. R. J. Mol Catal A: Chem. 2004, 213, 21-30 and references cited therein.  (7)  Grubbs, R. H . Tetrahedron 2004, 60,1111-1140 and references cited therein.  (8)  Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, ITh-lZZ and references cited therein.  Appendix A: SoHd-State Molecular Structures Detern,i„ed Dy A-Kay Crystallography  A.l Introduction The X-ray crystallographic data and the solid-state molecular structure solutions listed in this appendix have been collected and solved by the author. The compounds listed represent a broad cross-section of Legzdins-group chemistry over the past three years, and to place the structures in context reaction schemes and brief explanations have been included. The training and helpftil consultations with Brian Patrick over the years are gratefully acknowledged. His significant contributions to individual crystal solutions in terms of help with data collection and with solving twinned data-sets are noted where appropriate. Data collection for each compound was carried out at -100 ±1 °C on a Rigaku AFC7/ADSC C C D diffractometer or on a Bruker X8 A P E X diffractometer, using graphitemonochromated M o K a radiation. For each structure neutral-atom scattering factors were taken from Cromer and Waber.' Anomalous dispersión effects were included in Fcaic;' the valúes for A / and A / " were those of Creagh and McAuley.^ The valúes for mass attenuation coefficients are those of Creagh and Hubbell.'' A l l calculations were performed using the CrystalClear software package of Rigaku/MSC,^ or Shelxl-97.^  A.2 Products of Cp*W(NO)(CH2CMe3)2 with Cyclohexene in the Presence of Dihydrogen Xing (Michael) Jin reacted Cp*W(NO)(CH2CMe3)2 (1) with cyclohexene in the presence of dihydrogen to genérate three products, including A l and 43, as shown in Scheme A. 1. The solid-state molecular structure of 43 is presented in Chapter 5 of this thesis and that of A l is presented below. This system has been communicated.  Scheme A . l  42  Al  43  Figure A . l Solid-state molecular structure of A l with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-W(2) = 3.03146(18), W(l)-C(l) = 2.247(3), W(l)-N(l) = 1.773(3), N(l)-0(1) = 1.234(4), W(2)-C(7) = 2.242(3), W(2)-N(2) = 1.781(3), N(2)-0(2) = 1.229(4), W(l)-N(l)-0(1) = 167.3(3), W(2)-N(2)-0(2) = 166.6(3). The computed tungsten-hydride distances are as follows: W(l)-H(a) = 1.72, W(l)-H(b) = 1.83, W(2)H(a)= 1.83, W(2)-H(b)= 1.74. Data for A l were collected to a máximum 26» valué of 55.6 ° in 0.5 ° oscillations. The structure was solved by direct methods^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; Ha and Hb were refined isotropically, and all other hydrogen atoms were included in fixed posifions. Each bimetallic molecule was solvated by an EtiO molecule. The final cycle of fiíll-matrix least-squares analysis was based on 8657 observed reflections and 402 variable parameters. X-ray crystallographic data for the structure are presented in Table A . 1.  Table A . l X-ray Crystallographic Data for Complex A l . Al  Crystal Data Empirical formula  C35H64N2O3W2  Crystal Habit, color  Needle, red  Crystal size (mm)  0.50x0.15x0.08  Crystal system  Triclinic  Space group Volume (A^)  1863.29(5)  a (A)  9.23170(10)  b(A)  11.5165(2)  c(Á)  17.5796(3)  an  85.5750(10)  pn rn z  89.3080(10)  2  Density (calculated) (Mg/m^)  1.655  Absorption coefficient (mm'')  6.202  Fm  920  89.7330(10)  Data CoUection and Refinement Measured Reflections: Total  60049  Measured Reflections: Unique  8657  Final R índices'^  R l = 0.0210, wR2 = 0.0461  Goodness-of-fit onF^'' Largest diff peak and hole (e'  1.138 2.546 and -1.321  R l on F = S I (|Fol - |Fc|) i / E \F,l (/> 2a{I)); wR2 = [ (S ( F , ' -F,^f)IZ  w(Fo' f]"^ (all  data); w - [ G^F^^ ]"'; * GOF = [ S (w ( |Fo| - |Fc| f ) I degrees offreedom ]'^.  A.3 The Reactivity of Several Novel Cp*W(NO)(CH2CMe3)(allyl) Complexes Jenkins Tsang made, isolated and initiated reactivity studies on a variety of new Cp*W(NO)(CH2CMe3)(allyl) complexes. The first of these complexes, Cp*W(NO)(CH2CMe3)(CH2CHCHMe) (A2), forms an ii^-diene intermediate that can be trapped with PMe3 as A 3 , and effects the selective C - H activation of linear hydrocarbons to give complexes such as A4, A5, A 6 and A7, as shown in Scheme A.2. The crystallographic solutions for A5, and for A6/A7, which co-crystallized, model interesting and challenging disorder situations. Complexes A2, A 3 , A4 and A 5 have been published.^  Scheme A.2  A3 A2  A6  A7  Figure A.2 Solid-state molecular structure of A2 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.401(3), W(l)-C(2) = 2.346(3), W(l)-C(3) = 2.282(3), W(l)-C(5) = 2.257(3), W(l)-N(l) = 1.764(2), N(l)-0(1) = 1.221(3), C(l)-C(2) = 1.372(5), C(2)-C(3) = 1.425(4), C(3)-C(4) = 1.509(4), C(l)-C(2)-C(3) = 119.3(3), C(2)-C(3)-C(4) = 120.4(3), W(l)-C(5)-C(6) = 123.4(2), W(l)-N(l)-0(1) = 170.5(2). Data for A2 were coUected to a máximum 29 valué of 55.8 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l a , H l b , H2, H3, H5a and H5b were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of ñiU-matrix least-squares analysis was based on 4599 observed reflections and 232 variable parameters. X-ray crystallographic data for the structure are presented in Table A.2.  W1  Figure A.3 Solid-state molecular structure of A3 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.221(4), W(l)-C(2) = 2.218(3), W(l)-P(l) = 2.4335(8), W(l)-N(l) = 1.774(3), N(l)-0(1) = 1.230(4), C(l)-C(2) = 1.453(5), C(2)-C(3) = 1.456(5), C(3)-C(4) = 1.306(5), C(l)-C(2)-C(3) = 121.4(3), C(2)-C(3)C(4) = 126.7(4), W(l)-N(l)-0(1) = 170.1(3). Data for A3 were collected to a máximum 20 valué of 55.8 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l a , H l b , H2, H3, H4a and H4b were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 4503 observed reflections and 222 variable parameters. X-ray crystallographic data for the structure are presented in Table A.2.  W1  Figure A.4 Solid-state molecular structure of A4 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.333(4), W(l)-C(2) = 2.313(3), W(l)-C(3) = 2.294(3), W(l)-C(5) = 2.242(3), W(l)-N(l) = 1.788(3), N(l)-0(1) = 1.216(3), C(l)-C(2) = 1.363(5), C(2)-C(3) = 1.414(5), C(3)-C(4) = 1.501(5), C(l)-C(2)-C(3) = 118.8(3), C(2)-C(3)-C(4) = 120.6(3), W(l)-C(5)-C(6)= 116.5(2), W(l)-N(l)-0(1) = 174.9(2). Data for A4 were coUected to a máximum 29 valué of 56.4 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. The crystal was a two component twin. A l l non-hydrogen atoms were refined anisotropically; hydrogen atoms H l a , H l b , H2 and H3 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The fínal cycle of fiall-matrix least-squares analysis was based on 4679 observed reflections and 222 variable parameters. X-ray crystallographic data for the structure are presented in Table A.2.  Figure 5.1 Solid-state molecular structure of 41 with 50% probability thermal ellipsoids shown. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.269(4), W(l)-C(2) = 2.249(3), W(l)-C(7) = 2.388(4), W(l)-C(8) = 2.303(4), C(8)-C(7) = 1.392(6), C(7)-C(6) = 1.516(5), C(6)-C(l) = 1.517(5), C(l)-C(2) = 1.396(9), W(l)-N(l) = 1.787(3), N(l)-0(1) = 1.229(4), W(l)-N(l)-0(1) = 175.4(3), C(8)-C(7)-C(6) = 123.3(3), C(7)-C(6)-C(l) = 100.1(3), C(6)-C(l)-C(2)= 120.8(4).  To confirm that 41 is indeed a precursor to the catalytic cycle and not merely a byproduct, an isolated sample has been dissolved in cyclohexene and heated ovemight at 50 °C. Analysis of the final product mixture by ' H N M R spectroscopy reveáis the presence of cyclohexene oligomers and unreacted 41 (Figure 5.2), thereby indicating that it is indeed a precursor to the catalytically active species. Complex 41 can initiate oligomerization by dissociatíon of one of the olefinic linkages to open a coordination site at the metal center for the binding and coupling of cyclohexene or 1,4-cyclohexadiene. Evidence for this rafionale is provided by the E I - M S data of the 1,4-cyclohexadiene oligomer mixture which exhibits peaks  attributable to oligomers containing a vinyl end group (CH=CHPh) analogous to the 1,4-diene extant in 41 (Chapter 2.2.2).  «IB  I  «  •  ,  ,  6  ,  - ^ ^ ^ — p . . . ^  4  ,  ,  ^-~„..^—  s  5  —  Figure 5.2 ' H N M R spectrum (300 MHz, CeDe, rt) of the reaction mixture of 41 in neat cyclohexene after thermolysis at 50 °C for 16 h, demonstrating the formation of cyclohexene oligomers and the remaining presence of unreacted 41. The arrows indícate diagnostic peaks due to 41, while the cyclohexene oligomers give rise to peaks between 1 - 2 ppm and 5.3-5.8 ppm. As outlined in Chapter 2, the cyclohexene oligomers produced by precatalysts 5 and 6 are identical in form and distribution to those produced by precatalyst 1, with the exception of the small percentage of oligomers with capping end-groups. These end groups (ex. CH=CHPh or neopentyl) vary with the precatalyst employed and represent the initiation steps of the catalytic mechanism. This commonality in the products from different starting complexes suggests that variable initiation routes lead to a common catalytic species that is responsible for the cyclic olefín oligomerization. Examination of the precatalyst compounds shows that they all share the [Cp*W(NO)] core, with a variability in the organic ligands. Again, a catalytic species based on the [Cp*W(NO)] fragment is implied.  5.2.3 The proposed Cp*W(NO)(cyclic olefin)2 reactive species On the basis of comparisons between the tungsten precatalysts, their respective initiation reactions, and the molybdenum system, the [Cp*W(NO)] fragment is proposed as the basis for the catalytically active species. However, it is highly unlikely that this 14 electrón fragment would exist by itself in solution. Therefore, the solvated Cp* W(NO)(cyclic olefín)2 complex, as shown in Scheme 5.6, is proposed as a reasonable representation of the [Cp*W(NO)]-containing species. The coordinated olefins would then couple in the metal's coordination sphere to form a metallacyclopentane, a structural motif for which literature precedents exist.-'  Scheme 5.6  1  The coupled organic ligand of the tungstenacycle could undergo P-hydrogen eliminaüon to form a coordinated, mono-unsaturated cyclohexene dimer. On the simple tungstenacycle illustrated in Scheme 5.6 four p-hydrogens are available to the metal center, and due to the symmetry of the tungstenacycle, two possible dimer products could result. The newly opened space on the metal center would be filled by a third molecule of the cyclic-olefm substrate. The coordinated dimer can then be released, replaced at the metal center by a fourth olefin molecule to regenérate the Cp*W(NO)(cyclic olefin)2 complex. Altematively, the dimer can couple with the cyclic olefin to form a new tungstenacycle, and undergo P-hydrogen elimination to form a cyclohexene trimer. In an unsymmetrical metallacycle, such as occurs in the formation of trimers, up to four unique products might form at the elimination step. Scheme 5.7 summarizes the possible dimer and trimer products predicted to form from cyclohexene according to this proposed mechanism.  The experimentally observed cyclohexene oligomers correlate well with the general trends implicit in Scheme 5.7. As portrayed in the upper right of the scheme, two cyclohexene dimers are predicted, with the double bond in variable positions. The first of these predicted dimers matches the experimentally identified 3-cyclohexylcyclohexene (Chapter 2.2.7) which is the major mono-unsaturated dimer in the oligomer mixture. Up to twelve unique trimers are predicted by the proposed mechanism, which is strongly reminiscent of the family cluster made up of many trimer peaks observed in G C / M S analyses of the oligomers (Chapter 2, Figure 2.5). The ñirther complexity of the isolated cyclohexene oligomers (múltiple unsaturations) is accounted for by the previously established transfer dehydrogenation process (Chapter 2.2.3). The cyclohexene dimers and trimers capped with a neopentyl group result when the coupled neopentylcyclohexene from the initiation of precatalyst 1 remains coordinated to the metal center and subsequently couples to fiarther molecules of the cyclic-olefin substrate. When a similar process occurs with 4-methylcyclohexene as the cyclic substrate, peaks due to four unique neopentyl-capped dimers are evident in the GC/MS spectrum. The proposed mechanism suggests that these four compounds can be explained as the possible combinations for the position of the methyl groups relative to the neopentyl group. Thus, in the initial formation of the metallacycle that couples the neopentyl (in the form of the alkylidene) to the first 4methylcyclohexene molecule, the methyl group must be in one of two possible positions relative to the neopentyl, resulting in two products. In the subsequent coupling to the second molecule of 4-methylcyclohexene, the second methyl group also has two possible positions. In total, four combinations exist, leading to four peaks in the GC trace.  5.2.4 The proposed catalytic cycle for oligomerization of the cyclic olefíns Scheme 5.8 illustrates the proposed catalytic cycle for the oligomerization of cyclic olefins, beginning with the initiation pathways for precatalysts 1 and 5. The precatalysts form their respective reactive intermediates, couple an initial molecule of the cyclohexene substrate and lose the resultant organic ligand to form the common Cp*W(NO)(cyclic olefin)2 species. This species couples the two coordinated cyclohexene molecules in the metal's coordination sphere to form a tungstenacycle, which then undergoes p-hydrogen elimination to genérate a coordinated olefin. Finally, the coupled olefin may be replaced by cyclohexene to regenérate the  C p . W(NO)(cycl,c olefin), species. or aHematíve., teher couplings „ a y occur .o genérate h.gher-order oligomers of cyclohexene, illustrated Irere with one specific trimer example. Scheme 5.8  o  0^7  O o 24 Neopentyl coupled to cyclohexene  41  CH=CHPh coupled to cyclohexene  5.2.5 The catalytic cycle applied to the dimerization of allylbenzene by 3 The observed reactivity of 3 with the allylbenzene substrate to form 23 can be explained by a mechanistic pathway analogous to that just described for the cyclic olefms with precatalysts 1 and 5. Scheme 5.9 outlines the main steps, beginning with formation of 21 by reaction of one equivalent of allylbenzene with 3. The coupled diene ligand is released as 22, and two equivalents of allylbenzene coordínate to form a bis-olefm complex in which the steric interactions are minimized. Coupling of the olefms in the metal's coordination sphere produces the five-membered metallacycle. Then P-hydrogen activation and reductive elimination produce the observed dimer 23, and the active species is regenerated by coordination of fiírther allylbenzene. The specific p-hydrogen activation illustrated is likely favored by the benzylic nature of that hydrogen. This, together with the minimization of steric interactions in the bisolefin, accounts for the formation of a single, specific allylbenzene dimer."*  Scheme 5.9  3  5.2.6 Other examples of cyclic olefins coupled in the coordination sphere of Cp*W(NO) complexes The proposed mechanism for the coupling of cyclic olefins by a tungsten bis-olefin species outlined above in secüon 5.2.4 can be used to rationalize the observed reactivity of several other Legzdins systems for which mechanistic insight had been lacking. The first system involves the treatment of 1 with dihydrogen at rt, which results in the loss of neopentane to yield a putative Cp*W(NO)(CH2CMe3)(H) complex. When this reaction is carried out in cyclohexene as the substrate, three products are isolated, as illustrated in Scheme 5.10.-  Scheme 5.10  Product 42 has been previously synthesized by an altérnate route and is now identified by comparison of its characteristic spectral data.^ Products A l and 43 have been isolated and fully characterized, including X-ray crystallographic analyses. The solid-state structure of A l can be found in the Appendix of this thesis. The solid-state molecular structure of 43 (Figure 5.3) shows that two cyclohexene molecules have been coupled in the metal's coordination sphere, and it clearly displays the r|-',r|' binding motif of the resulting organic ligand. Compound 43 is proposed to form via a Cp*W(NO)(cyclohexene) intermediate (B), but no further mechanistic insight is offered in the original communication. A n isolated sample of 42 does not convert to 43  under the experimental conditions employed,^ so it cannot be invoked as an intermediate in the reaction mechanism. Also, use of D2 instead of H2 resuhs in no incorporation of deuterium in 42 and 43.  Figure 5.3 Solid-state molecular structure of 43 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.273(2), W(l)-C(8) = 2.284(2), W(l)-C(9) = 2.268(2), W(l)-C(10) = 2.434(2), W(l)-N(l) = 1.780(2), N(l)-0(1) = 1.225(3), C(l)-C(6) = 1.555(3), C(6)-C(7) = 1.559(3), C(7)-C(8) = 1.531(3), C(8)-C(9) = 1.413(3), C(9)-C(10) = 1.401(3), W(l)-C(l)-C(6) = 102.48(15), C(l)-C(6)-C(7) = 109.21(19), C(6)-C(7)-C(8)= 111.07(19), C(7)-C(8)-C(9) = 120.8(2), C(8)-C(9)-C(10) = 118.8(2), W(l)N(l)-0(1)= 174.81(18).  The formation of 43 from B can now be rationahzed further, as illustrated in Scheme 5.11, by using the key intermediates of the proposed thermolysis-induced oligomerization mechanism. Intermedíate B readily adds a second equivalent of cyclohexene to give the bisolefm adduct. After initial coupling to form the tungstenacycle, a double P-hydrogen activation results in the loss of H i and the formation of the observed ri^,!^'-allyl alkyl ligand.  Scheme 5.11  A single product (44), similar in structure to 43, is obtained from the hydrogenolysis of 1 in 1,3-cyclohexadiene at rt.^ A n X-ray crystallographic analysis reveáis a ri^,ri' binding motif in the organic ligand analogous to that found in 43, but there is an additional uncoordinated double bond in the ligand, as illustrated in Figure 5.4.  Figure 5.4 Solid-state molecular structure of 44 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.272(2), W(l)-C(8) = 2.295(2), W(l)-C(9) = 2.283(2), W(l)-C(10) = 2.424(2), W(l)-N(l) = 1.7817(17), N(l)-0(1) = 1.225(2), C(l)-C(2)= 1.502(3), C(2)-C(3) = 1.334(3), C(l)-C(6) = 1.558(3), C(6)-C(7) = 1.566(3), C(7)-C(8) = 1.531(3), C(8)-C(9) = 1.417(3), C(9)-C(10) = 1.399(3), W(l)-C(l)-C(6) = 103.66(13), C(l)-C(2)-C(3) = 125.0(2), C(l)-C(6)-C(7) = 109.11(16), C(6)-C(7)-C(8) = 109.86(18), C(7)-C(8)-C(9) = 120.5(2), C(8)-C(9)-C(10) = 118.9(2), W(l)-N(l)-0(1) = 173.66(17).  Scheme 5.12 illustrates the proposed mechanism by which 44 forms. Again, a bis-cyclic olefin complex which then undergoes coupling to form a tungstenacycle accounts for the observed product. In order to form 44 no loss of H2 is necessary since a slight shift coordinates a double bond beside one of the alkyl links to form an ri^-allyl and thus satisfies the electronic needs of the metal center.  Scheme 5.12  A comparison of the characteristic H N M R signáis for the three cyclohexene-derived products formed from 1 under hydrogenolysis conditions with the ' H N M R spectra observed for the product mixtures obtained from 1 under thermolysis conditions reveáis a startling result. None of the three hydrogenolysis products are present in the thermolysis mixture. Product A l would not be expected since it is not accessible via an alkylidene intermediate. But products 42 and 43, proposed to form via B, should be accessible under thermolysis conditions as well. Considering the case of 43 first, it seems that the temperature difference induces a change in reactivity after the formation of the tungstenacycle. As illustrated in Scheme 5.13, in the room temperature system P-hydrogen activation occurs twice, resulting in H i loss and formation of tiie stable allyl-alkyl product. Conversely, at 70 °C the initially activated P-hydrogen is transferred from the metal back to the ligand to form a coordinated monounsaturated dimer, and the complex continúes oligomerization. If, however, under thermolysis conditions a second Phydrogen activation beside the alkyl linkage is envisioned, dimers with two unsaturations could be formed with loss of dihydrogen. The resulting complexes could also retum to the oligomerization cycle with loss of the dimer. Recall that the molybdenum trans-metallacycles such as 9 lose H2 to form the t)"'-diene complexes.  í  1 in cyclohexene under H2 atm, rt  1 in cyclohexene, thermolysis at 70 °C  -2H  or  dimer, one unsaturation; reform [W](cyclohexene)2 dimer, two unsaturations; reform [W](cyclohexene)2  [W] = Cp*W(NO)  Considering 42, comparisons of several systems again suggest that temperature plays a key role. As seen above in Scheme 5.10, 42 forms at room temperature when 1 undergoes hydrogenolysis in cyclohexene, presumably via intermediate B (Scheme 5.14, fírst equation). Compound 42 was originally synthesized by the reaction of Cp*W(NO)(r|^CH2CHCMe2)(CH2CMe3) with cyclohexane at 50 °C (second equation).^ A mechanism involving múltiple C - H bond activations of the cyclohexane substrate, coupling of the allylderived fragment with the cyclohexane-fragment, and loss of that coupled ligand via C-H bond activation of a second equivalent of cyclohexane is proposed in the publication. The fínal intermediate is proposed to be the cyclohexene adduct (B) which undergoes a C - H bond activation to form 42.' Finally, thermolysis reactions of 1 in cyclohexane at 70 °C show that C - H activations again cause the formation of B, which is then trapped in the presence of PMes (third  equation). However, in the absence of a trapping agent, decomposition is observed at 70 °C instead of the isolation of 42 (fourth equation). Thus, it is clear that 42 is thermally stable to 50 °C for 6 h, but that the higher temperatures employed in the oligomerization reactions (70 °C for 40 h) lead to the decomposition of any 42 formed.  Scheme 5.14  A second Legzdins system has been reported to couple two molecules of a cyclic-olefin substrate under hydrogenolysis conditions. In this case, a general method was developed to react Cp*M(NO)(CH2SiMe3)2 ( M = M o or W) with H2 and acyclic dienes at low temperatures to lose two equivalents of tetramethylsilane and genérate r|''-trans-diene complexes. However, when  1,3-cyclooctadiene was employed as the diene, the resulting product (45) had two molecules of the substrate coupled in the metal's coordination sphere. The organic ligand was determined to be 2-cyclooct-2-en-l-yl-1,3-cyclooctadiene, a triene coordinated to the metal center in a bis-r)^ fashion, by single-crystal X-ray analysis. No mechanistic insight was possible at the time that these results were communicated.^ Now, application of the ideas of an initial bis-olefin complex followed by olefin-coupling in the metal's coordination sphere to form a metallacycle, p-hydrogen activation and rearrangement can rationalize this reactivity. Scheme 5.15 illustrates two possible routes by which the observed triene ligand may be formed.  Scheme 5.15  Finally, the mechanistic proposals that explain the cyclic-olefm oligomerization can rationalize the formation of 26, 27 and 28 during the thermolysis of 1 in cyclooctene, as well as the formation of cyclooctene dimers with one and two unsaturations (Chapter 4.2.3). Scheme 5.16 presents the proposed route to formation of 27 and 28. After the formation of the Cp*W(NO)(cyclic olefin)2 complex and subsequent conversión to the metallacycle, a double CH bond activation produces the 1,4-diene observed experimentally. Complex 27 in tum converts to 28, as shown by independent ' H N M R spectroscopy experiments (Chapter 4.2.3.2). Altematively, displacement of a cyclooctene dimer with either one or two unsaturations (double bonds) by the cyclooctene substrate reforms of the reactive Cp*W(NO)(cyclic olefin)2 complex.  Scheme 5.16  cyclooctene dimer, one unsaturation; and Cp*W(NO)(cyclic oIefin)2  cyclooctene dimer, two unsaturations; and Cp*W(NO)(cyclic olefin).  Compounds 27 and 28 are relatively stable with respect to temperature, and thus they can be isolated, whereas similar types of products are not observed or isolated when cyclohexene is the cyclic-olefm substrate. They can interconvert under thermolysis conditions, both in cyclic  olefín substrates and in CeDe. Finally, an isolated sample of 28 initiates fiírther oligomerization of cyclic olefíns under thermolysis conditions, demonstrating its ability to reenter the catalytic pathway. This likely occurs by loss of the diene ligand and coordination of two equivalents of the cyclic-olefin substrate to form the Cp*W(NO)(cyclic olefin)2 species, a process ftiUy in keeping with the general mechanism proposed in this chapter. As seen in the later steps of Scheme 5.13 and Scheme 5.16, the formation of a cyclicolefin dimer with two unsaturations requires the loss of two equivalents of hydrogen. At present insufficient data exists to state definitively in what form these hydrogens are lost. It may be as H2, as demonstrated by ' H N M R spectroscopy for the transformation from the trans-metallacycle to the ri'*-diene in the molybdenum system. Altematively, a direct metal-mediated transfer to the cyclic-olefin substrate may be occurring, as suggested by the ' H N M R data described in Chapter 2.2.3 for the cyclohexene and 1,4-cyclohexadiene oligomerizations with precatalyst 5. On the other hand, the transfer dehydrogenation process described in Chapter 2 may be facilitated by the catalytically active species in reaction steps that are independent of the steps of the oligomerization. At this point, no conclusions on the nature of the " 2 H " lost are supported by the available experimental data.  5.2.7 Comparison to historical cyclohexene oligomerization reactions The proposed catalytic cycle is based on a Cp*W(NO)(cyclic olefin)2 species which subsequentiy forms a tungstenacycle. In contrast, the 1970-80s oligomerizations of cyclohexene using tungsten-based metathesis catalysts are limited in their mechanistic insight.^ Only one report suggests a mechanism, said to involve a tungsten hydride species with which a cyclohexene molecule forms an adduct and subsequentiy reacts.^'' In contrast, the mechanisms proposed for the Legzdins systems described here do not contain any such species with both a coordinated cyclohexene molecule and a hydride. The transient hydride species proposed to form in the coupling of the cyclic olefins contain either an alkyl or an allyl linkage to the organic ligand. The stable hydride-containing compoimds which have been isolated (8, 28) are all allyl hydrides. Thus, hydride species analogous to those proposed in 1980 are neither observed ñor invoked in the present system.  5.2.8 Mechanistic proposals for the reaction of 1 and 5 with the oxygen-containing cyclic olefíns Oxygen-containing cyclic-olefin substrates tend to undergo ring-opening rather than oligomerization reactions with the tungsten precatalysts. Mechanistic explanations proposed here employ structures similar to those invoked for 1 and 5 with the cyclic olefíns, particularly in the preliminary formation of a metallacycle. Complex 1 reacts with the fíve-membered ring of 2,5-dihydrofuran to form a cismetallacycle (29). This is the expected product since in both the tungsten and molybdenum systems the small ring size forms a stable product and prevents ftirther reactivity. In contrast, 1 reacts with 3,4-dihydro-2H-pyran to form an alkoxy allyl complex in which the ring has been opened (30). As Scheme 5.17 proposes, the coupling of the alkylidene intermediate of 1 with the olefín forms the metallacycle. A P-hydrogen activation and transfer analogous to the cyclohexene system forms a coordinated olefín that also contains a metal-oxygen interaction. Cleavage of the C - 0 bond forms the alkoxy allyl complex observed.  Scheme 5.17  30  In a similar maimer, the reactivity of 5 with 3,4-dihydro-2H-pyran can be rationalized by coupling of the reactive ri^-alkyne intermediate with the substrate to form a metallacycle.  Subsequent (3-hydrogen activation and transfer forms a coordinated olefin whose heteroatom can interact with the metal center. The alkoxy allyl product 31 forms through C - 0 bond cleavage as shown in Scheme 5.18.  Scheme 5.18  5  31 In contrast, the reaction of 5 with 2,5-dihydrofiiran seems anomalous due to the atypical configuration observed in the final product 32. However, a speculative mechanism can be proposed. The expected initial step would be formation of a metallacycle through coupling of the intermediate species with the substrate. It is possible in this complex, as drawn in Scheme 5.19, for the oxygen atom to interact with the metal center, and thus become susceptible to a C - 0 bond cleavage and rearrangement that produces the observed product. A similarly-opened product is formed in an isoelectronic Cp2Zr(3-methoxybenzene) system, where the proposed mechanism is metallacycle formation by coupling of 2,5-dihydrofiiran with the 3-methoxybenzene, followed by C - 0 bond cleavage."' The difference in reactivity with precatalyst 1 is presumably due to the inability of the oxygen atom to approach the metal center because of the rigidity of the cismetallacyle.  5  \  32  5.2.9 Decomposition of the oligomerization catalyst: Concentration effects revisited As presented in Chapter 2, Table 2.4, the yield of cyclohexene oligomer produced by the tungsten precatalysts varies with the initial solution concentration. Decreasing the initial concentration of 1 causes an increase in the number of moles of cyclohexene converted per mole of precatalyst, with concentrations of 0.005 M - 0.010 M giving optimal conversions. Figure 5.5 illustrates the general trend for representative samples of 1 thermolyzed in cyclohexene for 40 h at 70 °C. The samples demónstrate the trend to higher substrate tumover at lower concentrations.  100  o  0.01  0.02  0.03  0.04  0.05  Concentration (M)  Figure 5.5 The effect of variable initial concentrations of 1 in cyclohexene on catalyst activity (tumover number, añer 70 °C, 40 h). The trend Une is drawn to guide the eye only, and error bars are calculated based on imcertainties of ± 2 mg of 1 and ± 1 0 mg of oligomer recovered.  The inverse trend observed suggests that lower concentrations discourage a catalyst decomposition pathway. Any oligomer-forming reaction would be expected to show a positive, proportional relationship between precatalyst levéis and total oligomer product formation, and the ratio of moles of oligomer formed per mole of precatalyst should not change in a given unit of time. The observed trend shows that another reaction (or perhaps several) is occurring that diverts the catalyst out of the catalytic cycle. If this catalyst decomposition occurred intramolecularly, the probability of its occurrence would be related to the relative rates of reaction, not the concentration of the rest of the solution. If the decomposition is intermolecular, with two tungsten-containing species coming together, then concentration dependence would be expected since increasing the concentration of the tungsten species would increase the probability of two such species meeting. The result would be a decrease in oligomer formation as the concentration of tungsten precatalyst increases, as is observed. The decomposition product(s) of the Cp* tungsten precatalysts have not been identified. However, for the Cp tungsten system, several decomposition products of 2 have been formulated as bimetallic species (Chapter 4.2.9.1). Compounds 36, 37, 38 and 39 are not likely candidates for a direct comparison to the Cp* system; however, they demónstrate the feasibility of a dimeric tungsten species. Similarly, compound A l in Scheme 5.10 shows the feasibility of a bimetallic tungsten species with Cp* ligands. A potential decomposition product in the oligomerization reactivity of 1 may be (Cp*W(N0))2.  The tungsten and molybdenum catalysts of Schrock also exhibit bimolecular decomposition pathways. Two modes of decomposition are observed in Schrock systems when ethylene is the metathesis substrate, although neither reaction has been elucidated in detall. First, a bimolecular coupling of the reactive alkylidene complex, particularly for methylene complexes, forms an ethylene complex that is no longer reactive to metathesis. Second, rearrangement of the metallocyclobutane intermediate leads to olefín formation by p-hydride elimination, and then to an ethylene complex identical to that formed by the fírst process. In the presence of substrates such as 2-pentene, bimetallic decomposition products are formed, those of the tungsten system having unsupported W=W bonds and those of the molybdenum system having bridging imido groups." In summary, both the experimental evidence and comparisons to related systems in the literature support the plausibility of a bimetallic decomposition pathway for the reactive  Cp*W(NO)(cyclic olefín)2 species and its derivatives. The number of operative pathways can not be determined, ñor can the details of the decomposition reaction(s) be elucidated. The nature of the final bimetallic species is unknown, although the (Cp*W(N0))2 complex is suggested.  5.3 Summary A mechanistic rationale for the catalytic oligomerization of cyclic olefms by thermolysis of precatalysts 1, 2, 5 and 6 in the olefín substrates has been presented. The initiation of precatalysts 1 and 5, and also of 3 in the dimerization of allylbenzene, involves the coupling of one equivalent of the substrate with the reactive intermediate typically formed by the precatalyst. The coupled ligand rearranges to an olefín, or to a diene with loss of H2, and is subsequently released from the metal center. The metal then coordinates two equivalents of substrate to form a bis-olefin complex. The bis-olefín formed from precatalysts 1 and 5, Cp*W(NO)(cyclic olefin)2, represents the convergent entry point to the catalytic cycle for both precatalysts. The coordinated olefins undergo metal-mediated coupling to form a metallacyclopentane complex. The metallacycle then undergoes P-hydrogen elimination and reductive elimination to genérate a cyclic-olefm dimer as a coordinated olefín. Further addition and coupling of substrate leads to formation of trimers and higher oligomers. Altematively, loss of the coordinated oligomer regenerates the reactive bis-olefm complex. The feasibility of this mechanism is supported by the available cyclic-olefm data from both the molybdenum and timgsten systems. The mechanism is fiírther supported by its ability to rationalize a range of couplings observed in Legzdins systems. The dimerization of allylbenzene by 3 yields a specific product (23) that is consistent with the proposed mechanism. The previously observed coupling of cyclohexene, 1,3-cyclohexadiene and 1,3-cyclooctadiene under hydrogenolysis conditions at ambient temperatures can now be explained as occurring via variations of the thermolysis mechanism. The isolated products from the thermolysis of 1 in cyclooctene also fit within the explanatory power of the proposed mechanism. Finally, decomposition for the tungsten catalyst species is consistent with a bimetallic pathway, the details of which have not been determined. The effect of this bimetallic decomposition is observed in the increase of cyclic-olefm conversión with decreasing precatalyst concentration.  5.4 Experimental Procedures  5.4.1 General Methods For applicable general methods, consuk section 4.4.1 in Chapter 4 of this thesis. Complex 41 was made by Dr. lan Blackmore. Details of its synthesis and characterization data are included here for completeness.  5.4.2 Preparation of 41 A re-sealable reaction vessel was charged with Cp*W(NO)(CH2SiMe3)(ri^-CPhCH2) (5) (350 mg, 0.65 mmol) and cyclohexene (8 mL). The red solution was then heated at 50 °C for 16 h, and the volatile components were removed from the final reaction mixture under reduced pressure to obtain an oily red solid. This solid was redissolved in a mínimum of pentane, and the solution was transferred to the top of a column of silica (ca. 6 x 0.5 cm). The organic products were eluted with a large volume of pentane, and the remaining red-brown band was eluted with 3:1 pentane/EtaO and the eluate was coUected. Removal of the volatile components from the eluate and re-crystallization of the residual solid from a mixture of Et20/hexanes at -30 °C afforded complex 41 as bright red crystals (93 mg, 0.17 mmol, 27%). Anal Caled for C24H31NOW ( 5 3 3 . 3 5 ) : C, 5 4 . 0 5 ; H , 5.86; N , 2 . 6 3 . Found: C, 54.14; H, 5.93; N , 2.91. IR (Nujol): VNO 1588 cm''. M S (EL 120 °C): m/z 535. ' H N M R (CeDg) 5; 7.20-7.18 (m, 3H, aryl), 6.98-6.92 (m, I H , aryl), 4.61 (m, I H , PhC//=CH), 3.14, 2.65, 1.72 and 1.96 (m, I H each, (overiapping) cycIohexenyl-C7^2), 2.22 (m, IH, olefinic CH), 1.83 (dd, I H , (overiapping) olefinic CH), 1.59 (m, 2H, (overiapping) cyclohexenyl-C//2), 1.41 (s, 1 5 H , CjMes), -0.97 (t, I H , olefinic-C//), cyclohexenyl-allylic protón obscured. '^C N M R (CeDg) 5; 146.7 (ipso-aryl), 128.6, 126.3, 124.7 (ortho, meta and para-ary/), 102.1 (CsMes), 69.2, 48.5, 45.2, 11.2 (olefinic carbons), 30.6 (cyclohexenyl-allylic carbón) 28.4, 27.7 17.7 (cyclohexenyl-CH2), 9.3 (CsA/es).  5.4.3 ' H N M R spectrum of 41 in cyclohexene after heating at 50 ° C for 16 hours A sample of 41 ( 1 0 mg) was dissolved in cyclohexene and heated at 5 0 °C for 16 h. The volatile components were removed in vacuo and the residue analyzed by ' H N M R spectroscopy.  5.4.4 X-ray Crystallography. Data collection was carried out at -100 ±1 °C on a Bruker X8 A P E X diffractometer, using graphite-monochromated Mo K a radiation. Data for 41 were collected to a máximum 19 valué of 55.8 ° in 0.5 ° oscillations. The structure was solved by direct methods'^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l , H2, H7 and H8 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 3957 observed reflections and 265 variable parameters. Data for 43 were collected to a máximum 19 valué of 57.0 ° in 0.5 ° oscillations. The structure was solved by direct methods'^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of fiall-matrix least-squares analysis was based on 4813 observed reflections and 231 variable parameters. Data for 44 were collected to a máximum 19 valué of 55.8 ° in 0.5 ° oscillations. The structure was solved by direct methods'^ and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H2, H3, H8, H9 and HIO were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 4467 observed reflections and 251 variable parameters. Neutral-atom scattering factors were taken from Cromer and Waber.Anomalous dispersión effects were included in Fcaic;'"* the valúes for A / and A / ' were those of Creagh and McAuley.'^ The valúes for mass attenuation coefficients are those of Creagh and Hubbell.'^ AU calculations were performed using S H E L X L - 9 7 . ' X - r a y crystallographic data for the three structures are presented in Table 5.1, and in the cif files.  Table 5.1 X-ray Crystallographic Data for Complexes 41, 43 and 44. 41  43  44  Empirical fornmla  C24H31NOW  C22H33NOW  C22H31NOW  Crystal Habit, color  Prism, red  Prism, red  Needle, pink  Crystal size (mm)  0.60 x 0.25 x 0.10  0.2x0.2x0.1  0.40 x 0.07 x 0.04  Crystal system  Orthorhombic  Monoclinic  Triclinic  Space group  Pna2]  C2/C  P-i  Volume (Á^)  2049.84(15)  3868.93(10)  948.71(14)  a (A)  16.8210(7)  22.1618(3)  8.1917(6)  biA)  12.6397(5)  13.0316(2)  9.7406(9)  c(A)  9.6412(4)  17.9373(3)  13.6903(13)  an  90  90  73.624(3)  j3n  90  131.682(2)  89.272(3)  rn z  90  90  65.724(3)  4  8  2  Density (calculated) (Mg/m^)  1.728  1.756  1.783  Absorption coefficient (mm'')  5.649  5.981  6.098  FDOO  1056  2032  504  Measured Reflections: Total  14882  66749  26325  Measured Reflections: Unique  3957  4813  4467  Final R índices'^  R l = 0.0178, wR2  R l = 0.0168, wR2  R l = 0.0152, wR2  = 0.0386  = 0.0379  = 0.0335  1.035  1.068  1.065  1.571 and -0.735  1.829 and -0.634  1.184 and -0.574  Crystal Data  Data Collection and Refinement  Goodness-of-fit on  *  Largest diff. peak and hole (e"  A-^) R l on F = E I (|Fo| - \FS | / E |Fo|, (/> 2a(7)); wR2 = [ (E ( Fo' - F,' fj/Z  w(F,' ff'  data); w = [ cr^Fo^ ]•'; * GOF = [ E (w ( |Fo| - \F,\ f ) / degrees offreedom ] ^'^.  (all  5.5 References (1)  Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, 223-233 and references therein.  (2)  Debad, J. D.; Legzdins, P.; Lumb, S. A.; Rettig, S. J.; Batchelor, R. J.; Einstein, F. W. B. Organometallics 1999,18, 3414 and references cited therein.  (3)  a) for proposals of intermediate metallacyclopentanes, see McLain, S. J.; Schrock, R. R. J. Am. Chem. Soc. 1978, i 1 3 1 5 - 1 3 1 7 and You, Y . ; Wilson, S. R.; Girolani, G . S. Organometallics 1994,13, 4655-4657. b) for examples of isolable metallacyclopentanes, see Ison, E. A.; Abboud, K . A.; Boncella, J. M . Organometallics 2006, 25, 1557-1564 and Amdt, S.; Schrock, R. R.; Müller, P. Organometallics 2007, 26, 1279-1290.  (4)  Graham, P. M . ; Buschhaus, M . S. A.; Pamplin, C. B.; Legzdins, P. Organometallics, 2008,27,2840-2851.  (5)  Jin, X . ; Legzdins, P.; Buschhaus, M . S. A. J. Am. Chem. Soc. 2005,127, 6928-6929.  (6)  Ng, S. H . K.; Adams, C. S.; Hayton, T. W.; Legzdins, P.; Patrick, B. O. J. Am. Chem. Soc. 2003,125, 15210-15223.  (7)  Jin, X . ; Legzdins, P., unpublished results.  (8)  Debad, J. D.; Legzdins, P.; Young, M . A.; Batchelor, R. J.; Einstein, F. W. B. J. Am. Chem. Soc. 1993,115, 2051-2052.  (9)  a) Moulijn, J. A.; van de Nouland, B. M . React. Kinet. Catal. Lett. 1975, 3, 405-408. b) Giezynski, R.; Korda, A . J. Mol. Catal. 1980, 7, 349-354.  (10)  Cuny, G. D.; Buchwald, S. L . Organometallics 1991,10, 363-365.  (11)  Schrock, R. R.; Czekelius, C. Adv. Synth. Catal. 2007, 349, 55-77 and references cited therein.  (12)  SIR92: Altomare, A . ; Cascarano, M . ; Giacovazzo, C ; Guagliardi, A. J. Appl. Cryst. 1993, 26, 343.  (13)  Cromer, D. T.; Waber, J. T. International Tables for X-ray Crystallography; Kynoch Press: Birmingham, 1974; Vol. IV.  (14)  Ibers, J. A . ; Hamilton, W. C. Acta Crystallogr. 1964,17, 781-782.  (15)  Creagh, D. C ; McAuley, W. J. International Tables of X-ray Crystallography; Kluwer Academic Publishers: Boston, 1992; Vol. C.  (16)  Creagh, D. C ; Hubbell, J. H. International Tables for X-ray Crystallography; Kluwer Academic Publishers: Boston, 1992; Vol. C.  (17)  SHELXL97: Sheldrick, G. M . University of Gottingen, Germany, 1997.  Chapter 6. Thesis Summary and Future Dírections  6.1 Thesis summary This thesis presents the investigations carried out to elucidate the nature, extent, and mechanism of the cyclic-olefm oligomerization reactivity observed with a series of tungsten precatalysts, with particular focus on complexes Cp*W(NO)(CH2CMe3)2 (1) and Cp*W(NO)(CH2SiMe3)(Ti^-CPhCH2) (5). Under thermolysis conditions these tungsten precatalysts oligomerize simple cyclic olefíns, from cyclopentene to cyclooctene, into ringretaining oligomers as high as dodecamers (depending on the substrate) with remaining sites of unsaturation. Chapter 2 describes in detall the oligomeric products obtained when the cyclicolefin substrates are cyclohexene and 1,4-cyclohexadiene. Chapter 3 presents the reaction pathway of Cp*Mo(NO)(CH2CMe3)2 (3) with cyclic olefíns in order to compare it to the reactivity of 1, the tungsten analogue of 3. Chapter 4 explores the reactivity of precatalyst 1 with a range of cyclic olefíns, with an emphasis on the detectable and isolable organometallic complexes that give insight into the oligomerizaion mechanism. Finally, Chapter 5 culminates in the proposal of a catalytic cycle that provides a mechanistic explanation for the oligomerization reactivity described in Chapters 2-4 and also for a range of other cyclic-olefm coupling reactions previously observed in Legzdins group chemistry.  6.1.1 Precatalyst initiation The mechanism of the cyclic-olefm oligomerization reaction is composed of a distinct initiation pathway for each precatalyst, a catalytic cycle that produces the oligomers, and a decomposition reaction (or reactions) that destroys the active catalytic species. The initiation pathway of precatalyst 1 is elucidated through comparison to the related molybdenum complex, Cp*Mo(NO)(CH2CMe3)2 (3). Both compounds genérate reactive alkylidene intermediates of the form [Cp*M(N0)(=CHCMe3)] ( M = W or Mo) through hydrogen transfer and loss of neopentane. The alkylidene intermediates then react with a substrate cyclic-olefm molecule via a 2 + 2 addition to form a metallacyclobutane. In the molybdenum system, the cis-metallacycle derived from cyclopentene (7) can be isolated, and it subsequently forms a stable allyl hydride complex (8). The cis-metallacycles derived from larger ring sizes isomerize to trans-metallacycles (9,14,17), presumably via an allyl hydride intermediate analogous to 8. During the thermolysis of 1, analogous tungsten trans-metallacycles derived from cyclohexene and cyclooctene (24, 26) are detected by ' H N M R spectroscopy; however, the putative preceding cis-metallacycles have not been observed. The molybdenum  trans-metallacycles convert to ri'*-diene complexes with loss of dihydrogen (10,15,16), and, under thermolysis conditions, the coupled organic ligand will dissociate from the metal center and allow a small amount of cyclic-olefm oligomerization. In contrast, the tungsten transmetallacycles cannot be isolated since vmder the thermolysis conditions required to genérate the alkylidene the resulting metallacycle converts to a coordinated olefín by P-hydrogen transfer. Subsequent loss of the olefin and coordination of two equivalents of the cyclic substrate forms the proposed catalytic species Cp*W(NO)(cyclic olefin)2. Thus, as summarized in Scheme 5.1, the initiation of precatalyst 1 follows a reaction pathway analogous to that outlined for molybdenum complex 3 but then readily proceeds into the catalytic oligomerization of cyclic olefins.  Scheme 6.1  1  n=l-4  Initiation of precatalyst 5 begins with formation of the reactive r)^-alkyne intermediate [Cp*W(NO)(ri^-HC=CPh)] by hydrogen transfer and loss of tetramethylsilane. The ri'-alkyne couples with a molecule of the cyclic olefin to form a metallacycle which then converts to a 1,4diene complex, an example of which has been isolated (41). Loss of the 1,4-diene ligand and  coordination of two equivalents of the cyclic substrate form the Cp*W(NO)(cyclic olefm)2 complex (Scheme 6.2).  Scheme 6.2  6.1.2 The catalytic cycle The proposed Cp*W(NO)(cyclic olefm)2 complex represents the entry point into the catalytic cycle, as illustrated in Scheme 6.3. In the tungsten's coordination sphere the cyclic olefíns couple to form a metallacyclopentane. Subsequent P-hydrogen activation and transfer yields a dihapto cyclic-olefin dimer, which may then either couple to a third molecule of the substrate or be displaced to regenérate the initial bis-olefin complex. hi this way, ring-retaining oligomers of various lengths and configurations are obtained. Consistent with the mechanism presented, the major cyclohexene dimer in the oligomer mixture is 3-cyclohexylcyclohexene. The detection of cyclohexene oligomers capped with neopentyl or CH=CPh end groups indicates that the coupled olefins formed during initiation of precatalysts 1 and 5 particípate in the coupling reactions in the catalytic cycle. While none of the species in the catalytic cycle have been directly observed, the coupling of the cyclic olefins in the coordination sphere of the tungsten complex is fiírther substantiated by the isolation of the 1,4-diene complex 27 and the allyl hydride complex 28, in which two molecules of cyclooctene have been coupled together. Tumover frequencies for precatalysts 1, 5 and 6 range from 5.5 to 6.5 mol/h at concentrations of 0.01 M in cyclohexene and 100 °C over a 24 h reaction time. Precatalyst 1 requires a mínimum temperature of 70 °C in order to readily form the initial alkylidene intermediate; the higher temperature of 100 °C increases both tumover and tumover frequency for the oligomerization of cyclohexene.  1,5  The coupling mechanism proposed to explain the oligomerization of simple cyclic olefms by precatalysts 1 and 5 can also be extended to rationalize the formation of other coupled products obtained under thermolysis or hydrogenolysis conditions. Precatalyst 1 under dihydrogen atmosphere at rt in the presence of cyclohexene or 1,3-cyclohexadiene forms allylalkyl complexes.' Under similar conditions Cp*M(NO)(CH2SiMe3)2 ( M = W or Mo) in 1,3cyclooctadiene forms an r)''-diene compound.^ In both cases, invoking the formation of a bisolefm complex that undergoes coupling to form a metallacyclopentane, analogous to the steps of the catalytic cycle above, rationalizes the formation of the observed products. Compound 3 catalytically dimerizes allylbenzene to yield a single product,3 and this selectivity is explained by the formation of a bis-olefm complex that minimizes the steric interactions of the coordinated allylbenzene substrate, followed by coupling and subsequent reléase of the organic product.  6.1.3 Catalyst decomposition and substrate limitations The active tungsten oligomerization catalyst is proposed to decompose via a bimetallic reaction pathway. The activity of precatalyst 1 in cyclohexene is concentration dependent, with greater substrate tumover achieved at lower precatalyst concentrations, which can be rationalized by such a decomposition mechanism. Heteroatoms within the cyclic-olefin ring, such as ether and amine functionalities, are detrimental to the oligomerization pathway. Oxygen-containing 2,5-dihydroñiran and 3,4dihydro-2H-pyran react with 1 and 5 to form metallacycles in a marmer analogous to the initiation pathways described in 6.1.1 (yide supra), followed by preferential activation of the accessible C - 0 bonds by the metal center to form ring-opened alkoxy products (30, 31, 32). The N - H bond of 1,2,3,6-tetrahydro-pyridine is activated by the alkylidene intermediate of 1 to form an amido product (33). In addition, the presence of a methyl group on the cyclic olefin leads to competing reactions. Oligomerization is thus currently limited to simple cyclic olefins.  6.1.4 Comparison of the Legzdins alkylidene complexes (Cp*M(NO)(=CHCMe3)) to known olefín metathesis alkylidene complexes The oligomeric products obtained from thermolysis of the tungsten precatalysts in cyclic olefms are ring-retaining. There is no evidence for R O M P processes in the systems examined in this thesis. As a result of ring-coupling, many isomers of the trimers and longer oligomeric lengths are formed, as evidenced by the number of peaks detected in the G C / M S analyses. In addition, the cyclohexene oligomers contain higher amounts of unsaturation than any related cyclohexene oligomer described to date in the literature. The high levéis of unsaturation observed are due to transfer dehydrogenation between the oligomer chains and the cyclic substrates, presimiably mediated by a tungsten species. Two early olefin metathesis experiments conducted in 1975 and 1980 report the ringretaining oligomerization of cyclohexene. These experiments, based on WCU in conjunction with a variety of additives to form catalytically-active reaction "soups", give oligomers similar to those obtained with precatalysts 1 and 5, but with much lower levéis of unsaturation.'' The reports provide very little mechanistic insight, but later work demonstrates that olefín metathesis is not the mechanism. In contrast, the oligomerization reactions described in this thesis begin with discrete precatalyst complexes whose initiation pathways can be determined, and the proposed catalytic cycle is supported by a variety of observations including several isolated  organometallic complexes containing two substrate molecules coupled in the metal's coordination sphere. The higher unsaturation levéis observed in the oligomers obtained from 1 and 5 introduce possibilities for functionalization at the olefinic sites. However, i f utilized in an industrial context, these products will occupy only a specialized niche market. It is of interest that 1, which forms a reactive alkylidene intermediate upon thermolysis, produces oligomers of cyclic olefins such as cyclopentene and cyclooctene but shows no signs of ROMP activity. The crucial involvement of alkylidene complexes in olefin metathesis has been known for many years and is amply demonstrated in the metathesis catalysts of Schrock and Grubbs.^"'' Comparisons between the group-six Schrock-alkylidene complexes and those of the Legzdins group reveal several key differences that ultimately result in very different modes of reactivity. The Schrock alkylidenes can be isolated and fully characterized, and during metathesis reactions some of the alkylidene complexes propagating the polymerization can be observed by N M R spectroscopy. Metallacyclobutane complexes are proposed as intermediate species but they are generally not detectable.^ In contrast, the Legzdins alkylidenes derived from 1 and 3 under thermolysis conditions are the undetectable, highly reactive intermediates* and the metallacycles resulting from coupling with cyclic olefins are the isolable or spectroscopically detectable species (vide supra). Both the Schrock and the Legzdins complexes are electrón defícient; however, the Schrock metathesis catalysts are formally d^, while the Legzdins precatalysts are formally á^. Thus, the Legzdins complexes are relatively electrón rich at the metal center, in large part because the nitrosyl ligand stabilizes the electrón rich d'' metal. These differences give the Legzdins alkylidene complexes their unique reactivity.  6.1.5 Signifícance and impact The research presented in this thesis provides unique insights into two major áreas; (1) the mechanism for formation of ring-retaining oligomers from cyclic olefins, with particular focus on cyclohexene and the tungsten nitrosyl precatalysts 1 and 5, and (2) the reactivity of the alkylidene intermediates formed from complexes 1 and 3 with cyclic olefins in a manner distinct from the R O M P reactivity commonly observed in other tungsten and molybdenum alkylidene systems. In terms of the advancement of Legzdins' group chemistry, this thesis contributes an extensive understanding of a new pathway of reactivity for a long-established compound. The CH bond activation ability of 1 has been known for some time; now its reactivity with cyclic olefins has been described. In the process of doing so, the reactions of 2, 3, 4 and 5 with cyclic  olefins have also been explored. In addition, the isolation and characterization of oligomers expands the focus of the group to consider the importance of organic products. The mechanism proposed in this thesis provides a unifíed explanation for the observed precatalyst initiation reactions and the catalytic cyclic-olefin oligomerization.  6.2 Future directions Several lines of experimentation can be envisioned for ñiture exploration, including an expansión of the possible substrates, further refinement of mechanistic understanding, and reduction of catalyst decomposition. Currently the substrates for oligomerization compatible with precatalysts 1 and 5 are limited to simple cyclic olefins. The larger ring sizes form relatively stable organometallic products compared to those of cyclohexene, and thus it will be increasingly difficult to effectively oligomerize substrates beyond cyclooctene. Ether and amine groups within the substrate ring have been shown to be detrimental to oligomerization. Further experiments should test a wider range of ñmctional groups for compatibility with the catalytic species. Ketones and polycyclic substrates might give interesting results. Substrates with readily activated C-H bonds will likely induce a competition between C-H bond activation and the desired oligomerization reactivity. Given the dimerization observed with 3, allylbenzene must be tested with the tungsten precatalysts despite the possibility of competitively activating the aryl hydrogen bonds. A selection of other acyclic olefins with available P-hydrogens could be tried as well. Mixed substrate systems could be explored, such as co-oligomerization of cyclohexene and cyclooctene. In terms of reaction condition optimization, the ideal concentration of precatalyst 1 in cyclohexene has been roughly determined. Further optimization should be done to refine the ideal reaction temperature and time conditions for both 1 and 5. Attempts to further probé the oligomerization mechanism in search of support (or lack thereof) for the proposed catalytic cycle will be difficult. The detection or the trapping of one of the species within the cycle would be ideal. More likely, the [Cp*W(NO)] fragment would be trapped, the trapping agent having displaced the coordinated cyclic olefins. The cholee of trapping agent will be of key importance since most conventional reagents such as phosphines will trap the alkylidene or ri'^-alkyne intermediates as they form. One potential possibility may  be 1,3-butadiene, which should react with the [Cp*W(NO)] fragment to form the previously synthesized, stable rj^'-trans-diene complex Cp*W(NO)(Ti''-CH2=CHCH=CH2).^ The 1,3butadiene may, however, couple to the intermediates formed during initiation in a marmer analogous to the reaction of 3 with acyclic olefms (Chapter 3, section 3.2.15). Kinetic studies will be hampered by the complexity of the reaction mixture and the uncertainty about the exact nature of the catalyst decomposition products. The catalyst lifetime might be extended by attaching the complex to some solid or polymeric support, such that the active species would be prevented from forming the bimetallic species proposed for the decomposition pathway. A reasonable site of attachment would be a tether replacing one of the methyl groups on the Cp* ligand of the precatalyst. In fact, similar catalyst supports have been proposed to prevent bimolecular decomposition in metathesis systems.^'' A n additional benefit would be the easy separation of the catalyst matrix from the oligomeric products and remaining cyclic-olefm substrate. Of course, a change in the reaction environment of this magnitude will inevitably affect the oligomerization reactivity of interest, whether by increasing, decreasmg, halting, or entirely changing the observed outcome.  6.3 References (1)  Jin, X . ; Legzdins, P.; Buschhaus, M . S. A. J. Am. Chem. Soc. 2005,127, 6928-6929, and unpublished observations (Jin, X . ; Legzdins, P.).  (2)  Debad, J. D.; Legzdins, P.; Young, M . A.; Batchelor, R. J.; Einstein, F. W. B. J. Am. Chem. Soc. 1993,115, 2051-2052.  (3)  Graham, P. M . ; Buschhaus, M . S. A.; PampHn, C. B.; Legzdins, P. Organometallics, 2008,27,2840-2851.  (4)  a) Moulijn, J. A . ; van de Nouland, B. M . React. Kinet.Catal. Lett. 1975, 3, 405-408, and b) Giezynski, R.; Korda, A . J. Mol Catal 1980, 7, 349-354.  (5)  For general references see a) Olefin Metathesis and Metathesis Polymerization; Ivin, K . J.; Mol, J. C ; Academic Press, San Diego, 1997 and references cited therein, and b) Handbook of Metathesis; Grubbs, R. H., Ed.; Wiley-VCH, Weinheim, 2003, vols. 1-3: Catalyst Development (vol. 1); Applications in Organic Synthesis (vol. 2); Applications in Polymer Synthesis (vol. 3) and references cited therein.  (6)  a) Schrock, R. R.; Czekelius, C. Adv. Synth. Catal 2007, 349, 55-77 and references cited therein. b) Schrock, R. R. J. Mol Catal A: Chem. 2004, 213, 21-30 and references cited therein.  (7)  Grubbs, R. H . Tetrahedron 2004, 60,1111-1140 and references cited therein.  (8)  Pamplin, C. B.; Legzdins, P. Acc. Chem. Res. 2003, 36, 223-233 and references cited therein.  Appendix A: SoHd-State Molecular Structures Determined Dy A-Kay Crystallography  A.l Introduction The X-ray crystallographic data and the solid-state molecular structure solutions Usted in this appendix have been coUected and solved by the author. The compounds Usted represent a broad cross-section of Legzdins-group chemistry over the past three years, and to place the structures in context reaction schemes and brief explanations have been included. The training and helpful consultations with Brian Patrick over the years are gratefuUy acknowledged. His significant contributions to individual crystal solutions in terms of help with data coUection and with solving twinned data-sets are noted where appropriate. Data coUection for each compound was carried out at -100 ±1 °C on a Rigaku AFC7/ADSC C C D diffractometer or on a Bruker X8 A P E X diffractometer, using graphitemonochromated M o K a radiation. For each structure neutral-atom scattering factors were taken from Cromer and Waber.' Anomalous dispersión effects were included in Fcaic;^ the valúes for A / and A / " were those of Creagh and McAuley.^ The valúes for mass attenuation coefficients are those of Creagh and Hubbell.'' AU calculations were performed using the CrystalClear software package of Rigaku/MSC,^ or Shelxl-97.^  A.2 Products of Cp*W(NO)(CH2CMe3)2 with Cyclohexene in the Presence of Dihydrogen Xing (Michael) Jin reacted Cp*W(NO)(CH2CMe3)2 (1) with cyclohexene in the presence of dihydrogen to genérate three products, including A l and 43, as shown in Scheme A. 1. The solid-state molecular structure of 43 is presented in Chapter 5 of this thesis and that of A l is presented below. This system has been communicated.  Scheme A . l  42  Al  43  Figure A . l Solid-state molecular structure of A l with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-W(2) = 3.03146(18), W(l)-C(l) = 2.247(3), W(l)-N(l) = 1.773(3), N(l)-0(1) = 1.234(4), W(2)-C(7) = 2.242(3), W(2)-N(2) = 1.781(3), N(2)-0(2) = 1.229(4), W(l)-N(l)-0(1) = 167.3(3), W(2)-N(2)-0(2) = 166.6(3). The computed tungsten-hydride distances are as follows: W(l)-H(a) = 1.72, W(l)-H(b) = 1.83, W(2)H(a)= 1.83, W(2)-H(b)= 1.74. Data for A l were coUected to a máximum 26» valué of 55.6 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; Ha and Hb were refined isotropically, and all other hydrogen atoms were included in fixed posifions. Each bimetallic molecule was solvated by an EtaO molecule. The final cycle of ñiU-matrix least-squares analysis was based on 8657 observed reflections and 402 variable parameters. X-ray crystallographic data for the structure are presented in Table A . 1.  Table A . l X-ray Crystallographic Data for Complex A l . Al  Crystal Data Empirical formula  C35H64N2O3W2  Crystal Habit, color  Needle, red  Crystal size (mm)  0.50x0.15x0.08  Crystal system  Triclinic  Space group Volume (Á^)  1863.29(5)  a (A)  9.23170(10)  b(A)  11.5165(2)  c(Á)  17.5796(3)  an  85.5750(10)  pn  89.3080(10) 89.7330(10)  z  2  Density (calculated) (Mg/m^)  1.655  Absorption coefficient (mm'')  6.202  Fm  920  Data Collection and Refinement Measured Reflections: Total  60049  Measured Reflections: Unique  8657  Final R índices'^  R l = 0.0210, wR2 = 0.0461  Goodness-of-fit onF^'' Largest diff peak and hole (e"  1.138 2.546 and -1.321  R l on F = S I (|Fol - |Fc|) i / E \F,l (/> 2a{I)); wR2 = [ (S ( F , ' -F,^f)IZ data); w = [ G^F^^ ]"'; * GOF = [ S (w ( |Fo| - |Fc| f ) I degrees offreedom  w(Fo' f]"^ (all .  A.3 The Reactivity of Several Novel Cp*W(NO)(CH2CMe3)(allyl) Complexes Jenkins Tsang made, isolated and initiated reactivity studies on a variety of new Cp*W(NO)(CH2CMe3)(allyl) complexes. The first of these complexes, Cp*W(NO)(CH2CMe3)(CH2CHCHMe) (A2), forms an ii^-diene intermediate that can be trapped with PMe3 as A 3 , and effects the selective C - H activation of linear hydrocarbons to give complexes such as A4, A5, A 6 and A7, as shown in Scheme A.2. The crystallographic solutions for A5, and for A6/A7, which co-crystallized, model interesting and challenging disorder situations. Complexes A2, A 3 , A4 and A 5 have been published.^  Scheme A.2  A3 A2  A6  A7  Figure A.2 Solid-state molecular structure of A2 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.401(3), W(l)-C(2) = 2.346(3), W(l)-C(3) = 2.282(3), W(l)-C(5) = 2.257(3), W(l)-N(l) = 1.764(2), N(l)-0(1) = 1.221(3), C(l)-C(2) = 1.372(5), C(2)-C(3) = 1.425(4), C(3)-C(4) = 1.509(4), C(l)-C(2)-C(3) = 119.3(3), C(2)-C(3)-C(4) = 120.4(3), W(l)-C(5)-C(6) = 123.4(2), W(l)-N(l)-0(1) = 170.5(2). Data for A2 were coUected to a máximum 29 valué of 55.8 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l a , H l b , H2, H3, H5a and H5b were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 4599 observed reflections and 232 variable parameters. X-ray crystallographic data for the structure are presented in Table A.2.  W1  Figure A.3 Solid-state molecular structure of A3 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.221(4), W(l)-C(2) = 2.218(3), W(l)-P(l) = 2.4335(8), W(l)-N(l) = 1.774(3), N(l)-0(1) = 1.230(4), C(l)-C(2) = 1.453(5), C(2)-C(3) = 1.456(5), C(3)-C(4) = 1.306(5), C(l)-C(2)-C(3) = 121.4(3), C(2)-C(3)C(4) = 126.7(4), W(l)-N(l)-0(1) = 170.1(3). Data for A3 were collected to a máximum 29 valué of 55.8 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l a , H l b , H2, H3, H4a and H4b were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 4503 observed reflections and 222 variable parameters. X-ray crystallographic data for the structure are presented in Table A.2.  W1  Figure A.4 Solid-state molecular structure of A4 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.333(4), W(l)-C(2) = 2.313(3), W(l)-C(3) = 2.294(3), W(l)-C(5) = 2.242(3), W(l)-N(l) = 1.788(3), N(l)-0(1) = 1.216(3), C(l)-C(2) = 1.363(5), C(2)-C(3) = 1.414(5), C(3)-C(4) = 1.501(5), C(l)-C(2)-C(3) = 118.8(3), C(2)-C(3)-C(4) = 120.6(3), W(l)-C(5)-C(6)= 116.5(2), W(l)-N(l)-0(1) = 174.9(2). Data for A4 were coUected to a máximum 29 valué of 56.4 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. The crystal was a two component twin. A l l non-hydrogen atoms were refined anisotropically; hydrogen atoms H l a , H l b , H2 and H3 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The fínal cycle of fiall-matrix least-squares analysis was based on 4679 observed reflections and 222 variable parameters. X-ray crystallographic data for the structure are presented in Table A.2.  Table A.2 X-ray Crystallographic Data for Complexes A2, A3 and A4. A2  A3  A4  Empirical formula  C19H33NOW  C17H30NOPW  C19H33NOW  Crystal Habit, color  Prism, yellow  Prism, yellow  Rod, yellow  Crystal size (mm)  0.45 X 0.30 X 0.20  0.35 X 0.20 X 0.20  0.35 X 0.20 X 0.10  Crystal system  Monoclinic  Monoclinic  Monoclinic  Space group  P2i/c  P2i/n  P2i/n  Volume (A^)  1940.63(8)  1875.15(16)  1919.05(17)  a (A)  9.0412(2)  8.1936(4)  8.9209(5)  b{A)  14.2430(3)  15.3514(8)  18.9252(9)  c(A)  15.6092(4)  14.9078(7)  11.3669(6)  a O  90  90  90  105.103(1)  90.020(1)  90.286(3)  Crystal Data  rn z  90  90  90  4  4  4  Density (calculated) (Wlglvc^)  1.627  1.698  1.645  Absorption coefficient (mm"')  5.955  6.245  6.022  Fooo  944  944  944  Measured Reflections: Total  29663  15837  36501  Measured Reflections: Unique  4599  4503  4679  Final R índices'^  R l = 0.0178, wR2  R l = 0.0216, wR2  R l = 0.0232, wR2  = 0.0423  = 0.0536  = 0.0526  1.080  1.014  1.038  1.185 and-0.806  1.579 and -0.932  1.619 and-0.774  Data Collection and Refinement  Goodness-of-fit on  *  Largest diff. peak and hole (e'  " R l on F = E I (|Foi - \F,\) | / Z |Fo|, (/ > 2a(7)); wR2 = [ (Z ( F,^ -F,^f)li: data); w = [ crVo^ ]"'; * GOF = [ Z (w (jFol - |Fc| f) / degrees offreedom  w(Fo' ff" .  (all  W1  Figure A.5 Solid-state molecular structure of A5 Part A with 35 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(la) = 2.368(14), W ( l ) C(2a) = 2.285(8), W(l)-C(3a) = 2.174(12), W(l)-C(5a) = 2.332(11), W(l)-N(l) = 1.772(3), N(l)-0(1) = 1.218(4), C(la)-C(2a) = 1.320(17), C(2a)-C(3a) = 1.398(12), C(3a)-C(4a) = 1.556(16), C(5a)-C(6a) = 1.099(16), C(la)-C(2a)-C(3a) = 117.4(11), C(2a)-C(3a)-C(4a) = 93.8(9), W(l)-C(5a)-C(6a) = 133.6(11), W(l)-N(l)-0(1) = 169.3(3). Data for A5 were coUected to a máximum 29 valué of 55.4 ° in 0.5 ° osciUations. The Structure was solved by direct methods and expanded using Fourier techniques. The two (mirror-image) chiral isomers of the compound co-crystalized to produce a crystalographically averaged solution. This was modeled as two disordered components called Part A and Part B for the methyl-allyl ligand and the ethyl ligand. Each part had 50% occupancy. AU non-hydrogen atoms were refined anisotropically. Related bond distances in the two parts were constrained to the same length within a standard deviatíon of 0.02 for the following pairs: Cla-C2a and C l b C2b; C2a-C3a and C2b-C3b; C3a-C4a and C3b-C4b; C5a-C6a and C5b-C6b; W l - C 5 a and W l C5b; Cla-C3a and Clb-C3b; and C2a-C4a and C2b-C4b. A l l hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 3767 observed reflections and 235 variable parameters. X-ray crystallographic data for the structure are presented in Table A.3.  Figure A.6 Solid-state molecular structures of a) A6 and b) Al with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg) for A6: W(l)-C(l) = 2.294(13), W(l)-C(2) = 2.48(1), W(l)-C(3) = 2.59(1), W(l)-C(5) = 2.18(2), W(l)-N(l) = 1.736(13), N(l)-0(1) = 1.163(17), C(l)-C(2) = 1.343(17), C(2)-C(3) = 1.319(17), C(3)-C(4) = 1.371(19), C(l)-C(2)-C(3)= 117.0(14), C(2)-C(3)-C(4)= 120.7(15), W(l)-N(l)-0(1) = 168.6(16). Selected interatomic distances (A) and angles (deg) for A7: W ( l ) - C ( l ) = 2.294(13), W(l)-C(2) = 2.48(1), W(l)-C(3) = 2.59(1), W(l)-C(6) = 2.134(15), W(l)-N(2) = 1.812(15), N(2)-0(2) = 1.23(2), C(l)-C(2) = 1.343(17), C(2)-C(3) = 1.319(17), C(3)-C(4) = 1.371(19), C(l)-C(2)-C(3) = 117.0(14), C(2)-C(3)-C(4) = 120.7(15), W(l)-N(2)-0(2) = 166.0(19). Data for A6/A7 were coUected to a máximum 2d valué of 61.6 ° in 0.5 ° osciUations. The structure was solved by direct methods and expanded using Fourier techniques. The two compounds co-crystallized in a 1:1 ratio. The space group imposed a mirror plañe on the crystalographically-averaged solution, effectively doubling each compound. The solution was therefore modeled as two disordered components called Part A and Part B for the methyl ligands and the nitrosyl ligands of A6 and A 7 respectively, and each part was modeled with 25% occupancy. The methyl-allyl ligands of A6 and A 7 overlapped chrystalographically, and were modeled with 50% occupancy. The Cp* ligands of the two compounds also overlapped, but imperfectly, causing distortion in the ellipsoid shapes. A l l non-hydrogen atoms were refined anisotropically, except N2, C5 and C6 which were refined isotropically. Related bond distances in the two parts were constrained to the same length within a standard deviation of 0.02 for the following pairs: W l - N l and W1-N2; N l - 0 1 andN2-02; and W1-C5 and W1-C6. A l l hydrogen  atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 1942 observed reflections and 136 variable parameters. X-ray crystallographic data for the structure are presented in Table A.3.  Figure A.7 Solid-state molecular structure of A6/A7 with 50 % probability thermal ellipsoids, showing the crystallographic mirror-plane and the relative positions of Part A (A6; N I , 01 and C5), Part B (A7; N2, 0 2 and C6), and the shared methyl-allyl ligand. One methyl group is removed from the Cp* ligand for clarity.  Table A.3 X-ray Crystallographic Data for Complexes A 5 and A6/A7. A5  A6/A7  Empirical formula  C,6H27NOW  C7.5H12.5N05O0.5Wc  Crystal Habit, color  Prism, yellow  Prism, orange  Crystal size (mm)  0.35 X 0.30 X 0.05 0.50 X 0.18 X 0.08  Crystal system  Monoclinic  Orthorhombic  Space group  F2i/n  Pcmb  Volume (Á^)  1622.3(4)  1541.1(13)  a (A)  11.7746(16)  7.085(5)  b(A)  9.4294(13)  15.437(5)  c(Á)  14.829(2)  14.091(5)  an  90  90  99.830(7)  90  rO  90  90  z  4  8  Density (calculated) (\Aglvci)  1.774  1.807  Absorption coefficient (mm"')  7.114  7.485  ^000  848  816  Measured Reflections: Total  21587  14198  Measured Reflections: Unique  3767  1942  Final R índices"  R l =0.0211, wR2  R l = 0.0258, wR2  = 0.0498  = 0.0611  1.064  1.117  Crystal Data  Data Collection and Refinement  Goodness-of-fit on  *  Largest diff peak and hole (e" 0.959 and -0.830  1.640 and-1.040  R l o n F = 11 (|Fo| - |Fc|) | / E \Fl (I> lail)); wR2 = [ (E (F,' -F.^f)/! data); w = [ cr^Fo^ ]"'; * GOF = [ E (w (|Fo| - |Fc| f ) / degi-ees offreedom  w(Fo' .  f](all  Complex A2 selectively activates the terminal methyl group of toluene to yield A 8 and of 2,3-dimethylbutene to yield A9 (Scheme A.3). In contrast, it couples to accessible double bonds to yield complexes such as A l O , A l l and A12 (Scheme A.4). Structures AS through A l l have been published.^'' Scheme A.3  All  W1  Figure A.8 Solid-state molecular structure of A8 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.410(7), W(l)-C(2) = 2.311(6), W(l)-C(3) = 2.299(6), W(l)-C(5) = 2.216(7), W(l)-N(l) = 1.770(5), N(l)-0(1) = 1.190(6), C(l)-C(2) = 1.395(9), C(2)-C(3) = 1.436(9), C(3)-C(4) = 1.458(10), C(l)-C(2)-C(3) = 120.5(6), C(2)-C(3)-C(4) = 120.5(6), W(l)-C(5)-C(6) = 117.8(4), W(l)-N(l)-0(1) = 170.2(5). Data for A8 were coUected to a máximum 26 valué of 56.0 ° in 0.5 ° osciUations. The structure was solved by direct methods and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of ñiU-matrix least-squares analysis was based on 4538 observed reflections and 223 variable parameters. X-ray crystallographic data for the structure are presented in Table A.4.  Figure A.9 Solid-state molecular structure of A9 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.361(6), W(l)-C(2) = 2.302(6), W(l)-C(3) = 2.290(6), W(l)-C(5) = 2.261(5), W(l)-N(l) = 1.761(6), N(l)-0(1) = 1.224(7), C(l)-C(2) = 1.357(9), C(2)-C(3) = 1.417(8), C(3)-C(4) = 1.485(10), C(5)-C(6) = 1.485(8), C(6)-C(7) = 1.350(8), C(l)-C(2)-C(3) = 120.4(7), C(2)-C(3)-C(4) = 120.6(7), W ( l ) C(5)-C(6) = 121.6(4), C(5)-C(6)-C(7) = 123.7(6), C(6)-C(7)-C(8) = 124.3(6), W(l)-N(l)-0(1) = 170.9(5). Data for A 9 were collected to a máximum 26 valué of 56.0 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. In the crystal lattice A 9 was solvated with THF in a 2:1 ratio. The disordered THF molecule was modeled with constrained, isotropic atoms. A l l other non-hydrogen atoms were refined anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of fiall-matrix least-squares analysis was based on 5277 observed reflections and 237 variable parameters. X-ray crystallographic data for the structure are presented in Table A.4.  Table A.4 X-ray Crystallographic Data for Complexes A8 and A9. A8  A9  C21H29NOW  C20H33NOW  Crystal Data Empirical formula  • !/2 (OC4H8) Crystal Habit, color  Prism, yellow  Irregular, orange  Crystal size (mm)  0.20 X 0.20 X 0.05  0.40 X 0.30 X 0.30  Crystal system  Orthorhombic  Monoclinic  Space group  F2,2i2i  C2/C  Volume (Á^)  1902.05(12)  4391.3(7)  a (A)  8.7093(3)  30.129(3)  b(A)  13.8379(5)  9.0268(9)  CiA)  15.7822(6)  17.0499(13)  aC)  90  90  Pi°) ri°)  90  108.735(3)  90  90  z  4  8  Density (calculated) (Mg/m^)  1.730  1.583  Absorption coefficient (mm"')  6.080  5.273  Fooo  976  2096  Measured Reflections: Total  17222  25263  Measured Reflections: Unique  4538  5277  Final R índices''  R l = 0.0273, wR2  R l = 0.0322, wR2  = 0.0581  = 0.0827  1.087  1.050  0.931 and-0.877  2.099 and-1.051  Data CoUection and Refinement  Goodness-of-fit on  *  Largest diff peak and hole (e'  R l on F = S I (|Fo| - IFcl) | / S \F¿ (/> 2a(/)); wR2 = [ (S ( Fo' - F^' )^ ) / 1 w(Fo' ff^ data); w = [ a^F¿^ ]"'; * GOF = [ 2 (w (|Fo| - \F,\ )^ ) I degrees offreedom  .  (all  Figure A.IO Solid-state molecular structure of A l O with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.227(4), W(l)-C(2) = 2.336(3), W(l)-C(3) = 2.437(4), W(l)-0(2) = 2.012(3), W(l)-N(l) = 1.763(3), N(l)-0(1) = 1.230(4), C(l)-C(2) = 1.447(6), C(2)-C(3) = 1.363(6), C(3)-C(4) = 1.488(5), C(4)-C(5) = 1.565(5), C(5)-0(2) = 1.413(4), C(l)-C(2)-C(3) = 117.9(4), C(2)-C(3)-C(4) = 124.3(4), C(3)C(4)-C(5) = 109.0(3), C(4)-C(5)-0(2) = 107.8(3), W(l)-0(2)-C(5) = 122.6(2), W(l)-N(l)-0(1) = 171.4(3). Data for A l O were collected to a máximum 20 valué of 56.0 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l a , H l b , H2 and H3 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 4493 observed reflections and 231 variable parameters. X-ray crystallographic data for the structure are presented in Table A.5.  C1  Figure A . l l Solid-state molecular structure of A l l with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.228(3), W(l)-C(2) = 2.299(3), W(l)-C(3) = 2.360(3), W(l)-C(6) = 2.191(3), W(l)-N(l) = 1.766(3), N(l)-0(1) = 1.212(4), C(l)-C(2)= 1.410(5), C(2)-C(3) = 1.381(5), C(3)-C(4) = 1.488(4), C(4)-C(5) = 1.503(4), C(5)-C(6) = 1.325(4), C(l)-C(2)-C(3) = 117.3(3), C(2)-C(3)-C(4) = 125.0(3), C(3)C(4)-C(5) = 108.9(2), C(4)-C(5)-C(6) =119.4(3), W(l)-C(6)-C(5) = 123.5(2), W(l)-N(l)-0(1) = 169.8(2). Data for A l l were coUected to a máximum 26 valué of 55.2 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l a , H l b , H2 and H3 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fiall-matríx least-squares analysis was based on 3834 observed reflections and 213 variable parameters. X-ray crystallographic data for the structure are presented in Table A.5.  C9 Figure A.12 Solid-state molecular structure of A12 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.267(19), W(l)-C(2) = 2.301(19), W(l)-C(3) = 2.44(2), W(l)-C(6) = 2.22(2), W(l)-N(l) = 1.78(2), N(l)-0(1) = 1.19(3), C(l)-C(2) = 1.43(3), C(2)-C(3) = 1.39(3), C(3)-C(4) = 1.49(3), C(4)-C(5) = 1.54(3), C(5)-C(6) = 1.46(3), C(l)-C(2)-C(3) = 120(2), C(2)-C(3)-C(4) = 127(2), C(3)-C(4)-C(5) = 104.3(19), C(4)-C(5)-C(6)= 110(2), W(l)-C(6)-C(5)= 115.9(16), W(l)-N(l)-0(1) = 170.6(17). Data for A12 were collected to a máximum 19 valué of 57.4 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. The crystal was twinned and one component of the twin was used to genérate the solution. (Brian Patrick helped to solve the twinning.) The crystallographic solution contained two molecules of A12. Atoms C6, C7, C14, C26, C31 and C35 were refined isotropically and all other non-hydrogen atoms were refined anisotropically. A l l hydrogen atoms were included in fixed positions. The final cycle of fiiU-matrix least-squares analysis was based on 8753 observed reflections and 396 variable parameters. The final solution was definite as to structure but unsuitable for publication due to the effects of the twinning on the R valúes. X-ray crystallographic data for the structure are presented in Table A.5.  Table A . 5 X-ray Crystallographic Data for Complexes A l O , A l l and A 1 2 . AlO  All  All  Empirical formula  C19H31NO2W  C18H27NOW  C20H31NOW  Crystal Habit, color  Píate, orange  Prism, orange  Rod, yellow  Crystal size (mm)  0.5 X 0.5 X 0.1  0.5 X 0.4 X 0.3  0.35 X 0.07 X 0.07  Crystal system  Monoclinic  Monoclinic  Triclinic  Space group  P2i/n  P2i/n  P-i  Volume (Á^)  1877.8(2)  1660.9(3)  1830.3(12)  a (A)  9.2125(8)  10.2949(9)  8.813(3)  b{k)  14.4236(12)  12.0055(13)  14.376(6)  c(A)  14.8434(3)  13.5878(14)  14.850(6)  an  90  90  103.120(10)  pn rn z  107.809(4)  98.512(5)  91.290(10)  90  90  92.040(10)  4  4  4  Density (calculated) (Mg/m^)  1.731  1.829  1.761  Absorption coefficient (mm"')  6.161  6.954  6.316  Fm  968  896  960  Measured Reflections: Total  19800  21396  8753  Measured Reflections: Unique  4493  3834  8753  Final R índices''  R l = 0.0256, wR2  R l = 0.0186, wR2  R l = 0.0906, wR2  = 0.0694  = 0.0464  = 0.2577  1.059  1.063  1.085  1.625 and-1.802  1.048 and-1.122  6.086 and -6.859  Crystal Data  Data CoUection and Refinement  Goodness-of-fit on  *  Largest diff peak and hole (e"  '^ R l on F = 11 (|Fo| - |Fc|) | / S |Fo|, (/> 2a(i)); wR2 = [ (S ( Fo' - F^^ )^ ) / 1 w(Fo' f]"^ (all data); w = [ aVo^ ]"'; * GOF = [ E (w (|Fo| - |Fc| f ) I degrees offreedom  .  As shown in Scheme A . 5 , complex A2 activates C-H bonds in the presence of an ether functional group (A13). It also activates tetramethylsilane to yield two products that cocrystallize (A14/A15), which in tum can also actívate C-H bonds in a marmer similar to A2. The structures of A14 and A l 5 have been published.'"  Scheme A.5  A14  Figure A,13 Solid-state molecular structure of A13 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.339(3), W(l)-C(2) = 2.320(3), W(l)-C(3) = 2.304(3), W(l)-C(5) = 2.245(3), W(l)-N(l) = 1.774(2), N(l)-0(1) = 1.223(3), C(l)-C(2) = 1.376(4), C(2)-C(3) = 1.413(4), C(3)-C(4) = 1.508(4), C(l)-C(2)-C(3) = 119.0(3), C(2)-C(3)-C(4) = 120.9(3), C(6)-C(5)-C(8) = 100.9(2), W(l)-C(5)-C(6) = 113.99(17), W(l)-C(5)-C(8) = 126.08(19), W(l)-N(l)-0(1) = 171.55(19). Data for A13 were collected to a máximum 20 valué of 56.0 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H l a , H l b , H2, H3 and H5 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of ñiU-matrix least-squares analysis was based on 4184 observed reflections and 225 variable parameters. X-ray crystallographic data for the structure are presented in Table A.6.  W1  Figure A.14 Solid-state molecular structure of A14 with 50 % probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(la) = 2.44(1), W(l)-C(2a) = 2.339(10), W(l)-C(3a) = 2.278(8), W(l)-C(5) = 2.212(6), W(l)-N(la) = 1.690(8), N(la)-0(la) = 1.210(11), C(la)-C(2a) = 1.392(16), C(2a)-C(3a) = 1.415(16), C(3a)-C(4a) = 1.504(14), C(la)C(2a)-C(3a) = 118.0(11), C(2a)-C(3a)-C(4a) = 119.3(10), W(l)-C(5)-Si(l) = 119.4(4), W ( l ) N(la)-0(la)= 167.5(10).  Figure A.15 Solid-state molecular structure of AIS with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(lb) = 2.35(3), W(l)-C(2b) = 2.39(2), W(l)-C(3b) = 2.41(3). W(l)-C(5) = 2.212(6), W(l)-N(lb) = 1.863(15), N(lb)-0(lb) = 1.275(17), C(lb)-C(2b) = 1.40(2), C(2b)-C(3b) = 1.41(2), C(3b)-C(4b) = 1.50(2), C(lb)-C(2b)C(3b) =118(3), C(2b)-C(3b)-C(4b) = 120(3), W(l)-C(5)-Si(l) = 119.4(4), W(l)-N(lb)-0(lb) = 171(2). Data for A14/A15 were coUected to a máximum 20 valué of 60.8 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A14 and A15 co-crystallized such that W l , C5, S i l and the Cp* ligand were crystalographically equivalent in both compounds. The rest of the solution was modeled as two disordered parts, Part A (A14) and Part B (A15), present in a 0.67 to 0.33 ratio. A l l non-hydrogen atoms were refined anisotropically, except for NIb, C6b and C3b which were refined isotropically. AU hydrogen atoms were included in fixed positions. The fínal cycle of fiíU-matrix least-squares analysis was based on 5957 observed reflections and 278 variable parameters. X-ray crystallographic data for the structure are presented in Table A.6.  Table A.6 X-ray Crystallographic Data for Complexes A13 and A14/A15. A13  A14/A15  Empirical formula  C18H29NO2W  C18H33NOSÍW  Crystal Habit, color  Prism, yellow  Prism, orange  Crystal size (mm)  0.14x0.10x0.09  0.8 X 0.5 X 0.5  Crystal system  Monoclinic  Monoclinic  Space group  P2,/n  P2x/c  Volume (A^)  1745.1(2)  1997.7(5)  a (A)  8.9259(7)  9.4739(14)  b(A)  14.7548(11)  13.991(2)  c(A)  13.3278(9)  15.631(2)  an  90  90  pn  96.183(3)  105.378(7)  90  90  z  4  4  Density (calculated) (Mg/m^)  1.809  1.634  Absorption coefficient (mm"')  6.627  5.845  -^000  936  976  Measured Reflections: Total  29919  33457  Measured Reflections: Unique  4184  5957  Final R Índices"  R l = 0.0187, wR2  R l = 0.0456, w  = 0.0366  = 0.1103  1.022  1.181  Crystal Data  Data Collection and Refinement  Goodness-of-fit on  *  Largest diff. peak and hole (e" 0.705 and -0.959  5.310 and -7.005  R l on F = E I (|Fo| - |Fc|) I / 1 \Fol (I > 2a(/)); wR2 = [ (S (F,^ - F,^ f ) 11. w{F,' f] data); w = [ a^F^^ ]"'; * GOF = [ S (w (|Fo| - \F,\ f) í degrees offreedom f'^.  (all  Ongoing attempts to induce fimctionalization of the alkyl ligand derived from C-H bond activation of a terminal methyl group have met with set-backs since the preferred cite of reactivity seems to be the allyl ligand. Thus, reaction of A14 with 2,6-dimethylphenyl isonitrile initially produces A l ó , which then isomerizes to A17 (Scheme A.6). Reaction of A14 with [CPh3][BF4] produces the organic product A18, which is derived from attack on the allyl ligand of A14 and subsequent reléase (Scheme A.7). A18 has the distinction of being the only purely organic solid-state molecular structure in this thesis.  Scheme A.6  A14  A18  Figure A.16 Solid-state molecular structure of A l ó Part B with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(5) = 2.072(8), W(l)N(2) = 2.156(6), W(l)-C(14) = 2.197(7), W(l)-N(l) = 1.756(6), N(l)-0(1) = 1.222(7), C(lb)C(2b) = 1.51(3), C(2b)-C(3b) = 1.29(2), C(3b)-C(4) = 1.55(2), C(4)-C(5) = 1.470(11), C(5)-N(2) = 1.257(10), N(2)-C(6) = 1.420(9), C(lb)-C(2b)-C(3b) = 124(2), C(2b)-C(3b)-C(4) = 122.7(18), C(3b)-C(4)-C(5) = 105.0(9), C(4)-C(5)-N(2) = 132.0(7), C(5)-N(2)-C(6) = 134.0(7), W(l)C(14)-Si(l) =118.1(3), W(l)-N(l)-0(1) = 172.3(6), C(lb)-C(2b)-C(3b)-C(4) = 177.7(17). C(2b)-C(3b)-C(4)-C(5) = 133.5(17), C(4)-C(5)-N(2)-C(6) = 8.0(15). Data for A l ó were collected to a máximum 20 valué of 56.4 ° in 0.5 ° oscillations. The Structure was solved by direct methods and expanded using Fourier techniques. The C1-C2-C3 fragment was disordered in two orientations, modeled as two parts (A and B) each with 50% occupancy. As a result of their proximity C3a and C3b were refined isotropically, while all other non-hydrogen atoms were refined anisotropically. A l l hydrogen atoms were included in fixed positions. The final cycle of fiall-matrix least-squares analysis was based on 6344 observed reflections and 318 variable parameters. X-ray crystallographic data for the strucmre are presented in Table A.7.  Figure A.17 Solid-state molecular structure of A l ? with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(5) = 2.0873(19), W(l)-N(2) = 2.1315(16), W(l)-C(14) = 2.211(2), W(l)-N(l) = 1.7616(16), N(l)-0(1) = 1.223(2), C(l)-C(2) = 1.482(5), C(2)-C(3) = 1.483(4), C(3)-C(4) = 1.319(4), C(4)-C(5) = 1.439(3), C(5)-N(2) = 1.260(2), N(2)-C(6) = 1.416(2), C(l)-C(2)-C(3) =113.2(3), C(2)-C(3)-C(4) = 125.9(3), C(3)C(4)-C(5) = 122.5(2), C(4)-C(5)-N(2) = 129.18(19), C(5)-N(2)-C(6) = 133.21(17), W(l)-C(14)Si(l) = 119.59(10), W(l)-N(l)-0(1) = 170.73(17), C(2)-C(3)-C(4)-C(5) = 175.7(3), C(4)-C(5)N(2)-C(6)= 12.5(4). Data for A17 were coUected to a máximum 20 valué of 61.4 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically. Hydrogens H3 and H4 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fiall-matrix least-squares analysis was based on 8712 observed reflections and 308 variable parameters. X ray crystallographic data for the structure are presented in Table A.7.  Figure A. 18 Solid-state molecular structure of A18 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): C(l)-C(2) = 1.492(2), C(2)-C(3) = 1.310(2), C(3)-C(4) = 1.502(2), C(4)-C(5) = 1.562(2), C(l)-C(2)-C(3) = 126.49(17), C(2)-C(3)-C(4) = 123.84(15), C(3)-C(4)-C(5) = 119.33(13), C(l)-C(2)-C(3)-C(4) = 177.01(16), C(2)-C(3)-C(4)C(5)= 158.46(16). Data for A18 were collected to a máximum 20 valué of 54.8 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. Due to equipment failure, only a partial data set was collected (80%) and therefore this structure cannot be published. Despite this, the connectivity and R valúes are sufficiently good to include it here. A l l non-hydrogen atoms were refined anisotropically. Hydrogens H2 and H3 were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 3260 observed reflections and 217 variable parameters. X-ray crystallographic data for the structure are presented in Table A.7.  Table A.7 X-ray Crystallographic Data for Complexes A l ó , A17 and A18. Aló  A17  AIS  Crystal Data Empirical formula  C27H42N2OSÍW  C27H42N2OSÍW  C23H22  Crystal Habit, color  Irregular, yellow  Prism, red-orange  Needle, yellow  Crystal size (mm)  0.15 x 0.125 x 0.075 0.50 X 0.40 X 0.25  0.45 X 0.30 X 0.20  Crystal system  Triclinic  Monoclinic  Monoclinic  Space group  P-i  P2i/n  P2,/c  Volume (Á^)  1338.0(4)  2821.1(4)  1672.13(17)  a (A)  9.8708(16)  9.7644(8)  16.0109(10)  b(A)  11.1838(17)  16.7802(14)  14.4336(9)  c(A)  12.928(2)  17.2285(15)  7.2497(4)  «O  84.215(8)  90  90  fin rn z  87.804(9)  91.998(4)  93.563(2)  70.445(8)  90  90  2  4  4  Density (calculated) (Mg/m^)  1.545  1.466  1.185  Absorption coefficient (mm'')  4.382  4.157  0.067  Fooo  628  1256  640  Measured Reflections: Total  35982  54722  9091  Measured Reflections: Unique  6344  8712  3260  Final R índices"  R l = 0.0421, wR2  R l = 0.0181, wR2  R l = 0.0404, wR2  = 0.1211  = 0.0419  = 0.1061  1.055  1.045  1.017  1.962 and-2.460  0.950 and -0.920  0.223 and -0.189  Data CoUection and Refinement  Goodness-of-fit on  *  Largest diff. peak and hole (e"  " R l on F = S I (|Fo| - \FS | / E |Fo|, ( / > 2am data); w = [ cr^Fo^ ]•';  *GOF= [ S  (w (  |Fo| -  wR2 = [ (S (F,^ -F,^f)/1 \F,\ f ) / degrees offreedom  w(Fo' ff^ .  (all  A second novel allyl complex Cp*W(NO)(CH2CMe3)(CH2CHCHPh) (A19) has been made by Jenkins Tsang. Scheme A.8 shows two of its reactions to produce allyl hydride products A20 and All. The structure of A19 has been published'*' but those of the products have not. Scheme A.8  A21 The novel allyl complexes Cp*W(NO)(CH2CMe3)(CH2CMeCH2) and Cp*W(NO)(CH2CMe3)(CH2CHCH2) have both been reacted with pyrrolidine to yield A22 and A23, respectively (Scheme A.9). The structures of A22 and A23 have been published."' Scheme A.9  W1  Figure A.19 Solid-state molecular structure of A19 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.397(3), W(l)-C(2) = 2.345(3), W(l)-C(3) = 2.304(3), W(l)-C(10) = 2.265(3), W(l)-N(l) = 1.772(3), N(l)-0(1) = 1.223(4), C(l)-C(2) = 1.365(5), C(2)-C(3) = 1.437(5), C(3)-C(4) = 1.494(4), C(l)-C(2)-C(3) = 120.9(3), C(2)-C(3)-C(4) = 121.0(3), W(l)-C(10)-C(l 1) = 123.6(2), W(l)-N(l)-0(1) = 169.7(2). Data for A19 were coUected to a máximum 26 valué of 55.8 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refmed anisotropically; hydrogen atoms H l a , H l b , H2 and H3 were refmed isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 5262 observed reflections and 268 variable parameters. X-ray crystallographic data for the structure are presented in Table A.8.  Figure A.20 Solid-state molecular structure of A20 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.262(3), W(l)-C(2) = 2.328(2), W(l)-C(3) = 2.360(2), W(l)-N(l) = 1.763(2), N(l)-0(1) = 1.216(3), C(l)-C(2) = 1.403(3), C(2)-C(3) = 1.403(3), C(3)-C(4) = 1.492(3), C(2)-C(10) = 1.521(3), C(10)-C(ll) = 1.492(3), C(ll)-C(12)= 1.317(4), C(l)-C(2)-C(3) = 119.0(2), C(2)-C(3)-C(4) = 125.6(2), C(l)C(2)-C(10) = 122.9(2), C(2)-C(10)-C(l 1) = 114.8(2), C(10)-C(l 1)-C(12) = 123.4(2), W(l)-N(l)0(1)= 169.4(2). Data for A20 were collected to a máximum 20 valué of 60.6 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms HOl (hydride), H l a , H l b , H3, HIO, H l 1 and H12 were refined isotropically, and all other hydrogen atoms were included in fixed posifions. The final cycle of full-matrix least-squares analysis was based on 6219 observed reflections and 286 variable parameters. X-ray crystallographic data for the structure are presented in Table A.8.  Figure A.21 Solid-state molecular structure of A21 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.422(10), W(l)-C(2) = 2.303(9), W(l)-C(3) = 2.278(9), W(l)-N(l) = 1.759(7), N(l)-0(1) = 1.228(10), C(l)-C(2) = 1.405(15), C(2)-C(3)= 1.368(15), C(3)-C(4) = 1.464(14), C(l)-C(10) = 1.450(15), C(l)-C(2)C(3) = 120.5(10), C(2)-C(3)-C(4) = 123.0(10), C(2)-C(l)-C(10) = 126.7(10), W(l)-N(l)-0(1) = 170.0(7). Data for A21 were coUected to a máximum 20 valué of 56.4 ° in 0.5 ° osciUations. The Q  structure was solved by direct methods and expanded using Fourier techniques. The crystal was a racemic twin, and the crystallographic solution had one solvent molecule (EtaO) per two A21 molecules. The hydride of the first A21 molecule could be modeled on the basis of the residual electrón density. The disordered nitrosyl ligand of the second A21 molecule was modeled in two positions and the hydride could not be modeled. Disorder in the EtaO molecule could not be fuUy modeled. A l l non-hydrogen atoms were refined anisotropically; hydrogen atom HOl (hydride) was refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 11491 observed reflections and 580 variable parameters. X-ray crystallographic data for the structure are presented in Table A.8.  Table A.8 X-ray Crystallographic Data for Complexes A19, A20 and A21. A19  A20  A21  C24H35NOW  C25H33NOW  C25H29NOW  Crystal Data Empirical formula  • '/2(C4H,oO)  Crystal Habit, color  Rod, yellow-orange  Prism, orange  Prism, yellow  Crystal size (mm)  1.4x0.1 x O . l  0.6 X 0.6 X 0.35  0.40x0.20 x 0.  Crystal system  Monoclinic  Triclinic  Orthorhombic  Space group  P2i/c  P.i  P2,2,2,  Volume (A^)  2223.15(17)  1066.2(3)  4812.5(10)  a (A)  12.7252(6)  9.4452(15)  9.3219(9)  biA)  11.7231(5)  11.0514(17)  22.502(3)  ciA)  15.2438(7)  11.6303(16)  22.943(3)  an  90  71.176(7)  90  102.146(2)  86.420(8)  90  90  68.403(8)  90  4  2  8  1.705  1.602  z Density (calculated)  (Mg/rn^)1.606  Absorption coefficient (mm"')  5.209  5.432  4.821  Fooo  1072  544  2312  Measured Reflections: Total  27675  31022  38299  Measured Reflections: Unique  5262  6219  11491  Final R índices"  R l = 0.0210, wR2  Rl  = 0.0660  wR2 = 0.0491  = 0.1327  1.153  1.102  1.064  1.504 and-1.371  0.860 and-1.351  3.729 and -2.287  Data Collection and Refínement  Goodness-of-fit on  *  Largest diff. peak and hole (e'  =  0.0204,  R l = 0.0530, wF  " R l on F = 11 (|Fo| - |Fe|) | / 1 |Fo|, ( / > 2CT(Í)); W R 2 = [ ( S ( F o ' - F / ) ' ) / S w(Fo' ff'^  data); w = [ a V o ^ ]"'; * GOF = [ S (w (|Fo| - |Fc| f ) I degrees offreedom  .  (all  Figure A.22 Solid-state molecular structure of A22 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.184(3), W(l)-N(2) = 1.938(2), W ( l ) - N ( l ) = 1.763(2), N(l)-0(1)= 1.239(3), C(ll)-C(12) = 1.324(5), C(10)-C(ll)C(12) = 121.6(4), W(l)-N(2)-C(9) = 127.17(18), W(l)-N(2)-C(6) = 124.65(18), W(l)-N(l)-0(1) = 168.7(2). Data for A22 were coUected to a máximum 20 valué of 55.0 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of fiíU-matrix least-squares analysis was based on 5392 observed reflections and 248 variable parameters. X-ray crystallographic data for the structure are presented in Table A.9.  C6 Figure A.23 Solid-state molecular structure of A23 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(2) = 2.198(6), W(l)-N(2) = 1.926(4), W(l)-N(3) = 1.749(5), N(3)-0(l) = 1.246(5), C(l)-C(2) = 1.537(7), C(2)-C(3) = 1.544(6), C(3)-N(l) = 1.462(7), C(l)-C(2)-C(3) = 110.0(4), C(2)-C(3)-N(l) = 114.3(5), C(3)N(l)-C(4) = 114.0(4), C(3)-N(l)-C(7) = 112.9(5), W(l)-N(2)-C(8) = 128.7(3), W(l)-N(2)-C(l 1) = 125.1(3), W(l)-N(3)-0(1) = 168.8(4). Data for A23 were coUected to a máximum 29 valué of 47.0 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. The crystal was twinned, and the crystallographic solution contained two molecules of A23. In the second molecule, atom C l O A was disordered in two positions in a 0.65 to 0.35 ratio and was modeled isotropically. A l l other non-hydrogen atoms were refmed anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of fiall-matrix least-squares analysis was based on 11672 observed reflecüons and 481 variable parameters. X-ray crystallographic data for the structure are presented in Table A.9.  Table A.9 X-ray Crystallographic Data for Complexes All and A23. A22  A23  Empirical formula  C23H40N2OW  C21H37N3OW  Crystal Habit, color  Prism, yellow  Needle, yellow  Crystal size (mm)  0.40x0.175 x 0.075  0.55 X 0.075 x 0.05  Crystal system  Monoclinic  Triclinic  Space group  Plx/c  P-x  Volume (Á^)  2347.98(14)  2217.0(6)  a (A)  12.6452(4)  8.0022(10)  b(A)  8.1081(3)  16.383(3)  c(A)  22.9995(8)  17.082(3)  an  90  84.451(6)  pn rn z  95.311(2)  89.925(6)  90  84.088(6)  4  4  Density (calculated) (Mg/m^)  1.540  1.592  Absorption coefficient (mm'')  4.934  5.224  ^000  1096  1064  Measured Reflections: Total  36666  11672  Measured Reflections: Unique  5392  11672  Final R índices"  R l = 0.0200, wR2  R l = 0.0315, wR2  = 0.0500  = 0.0553  1.061  0.883  1.731 and-0.674  1.925 and -0.957  Crystal Data  Data Collection and Refinement  Goodness-of-fit on  *  Largest diff peak and hole (e' Á'^)  " R l on F = S I (|Fo| - |Fc|) I / S \F¿ (I > IGÍP)); W R 2 = [ ( S ( F,^ -F,^f)IY.  w(Fo^ f]  data); w = [ O^FJ" ]''; * GOF = [ S (w (|Fol - \Fc\ f ) / degrees offreedom f"^.  (all  A.4 Peroxide-Induced Nitrosyl Insertion Peter Graham isolated compound A24, where the nitrosyl ligand of 3 has inserted into one of the neopentyl ligands and the incoming oxygen atom has trapped the resulting complex (Scheme A. 10).  Scheme A. 10  3  A24  A n analogous tungsten complex (A25) was subsequently made from Cp*W(NO)(CH2SiMe3)2 (Scheme A . l 1). This nitrosyl insertion is believed to be the initial step in the oxidation reaction that ultimately forms Cp*W(CH2SiMe3)(0)2 when dioxygen is the oxidant.  Scheme A . l l  Figure A.24 Solid-state molecular structure of A24 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): Mo(l)-N(l) = 2.039(3), Mo(l)-0(l) = 1.951(3), Mo(l)-0(2) = 1.699(3), Mo(l)-C(6) = 2.204(4), N(l)-0(1) = 1.417(4), N(l)-C(l) = 1.469(5), Mo(l)-N(l)-C(l) = 115.1(3), Mo(l)-N(l)-0(l) = 65.90(17), Mo(l)-0(l)-N(l) = 72.58(17), N(l)-Mo(l)-O(l) = 41.52(12). Data for A24 were collected to a máximum 10 valué of 50.4 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically, and all hydrogen atoms were included in fixed positions. 0 1 , 0 2 and N I were independently refined for occupancy to confirm their identity. The fínal cycle of fiíU-matrix least-squares analysis was based on 3839 observed reflections and 228 variable parameters. X-ray crystallographic data for the structure are presented in Table A.IO.  W1  Figure A.25 Solid-state molecular structure of A25 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-N(l) = 2.0360(14), W(l)-0(1) = 1.9475(13), W(l)-0(2) = 1.7075(14), W(l)-C(5) - 2.1555(19), N(l)-0(1) = 1.447(2), N(l)-C(l) = 1.482(2), W(l)-N(l)-C(l) = 117.37(12), W(l)-N(l)-0(1) = 65.46(7), W(l)-0(1)-N(l) = 72.00(8), N(l)-W(l)-0(1) = 42.54(6). Data for A25 were coUected to a máximum 10 valué of 61.2 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refmed anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 6894 observed reflections and 228 variable parameters. X-ray crystallographic data for the structure are presented in Table A . 10.  Table A.IO X-ray Crystallographic Data for Complex A24 and A25. A24  A25  Empirical forumla  C20H37M0NO2  C,8H37N02SÍ2W  Crystal Habit, color  Irregular, yellow  Prism, yellow  Crystal size (mm)  0,35 X 0.30 X 0.05 0.65 X 0.45 X 0.25  Crystal system  Triclinic  Triclinic  Space group  P-i  P.i  Volume (Á^)  1083.32(17)  1139.64(18)  a (A)  8.6671(8)  9.9833(9)  b(A)  10.2287(9)  10.6781(10)  c(A)  13.5257(12)  12.7497(11)  an  103.417(4)  110.998(4)  105.869(4)  111.235(4)  100.361(4)  95.911(4)  2  2  1.286  1.572  Absorption coefficient (mm"')  0.616  5.183  FQOO  444  540  Measured Reflections: Total  28878  33308  Measured Reflections: Unique  3839  6894  Final R índices"  R l = 0.0404, wR2  R l = 0.0172, wR2  = 0.1098  = 0.0412  1.051  1.121  1.464 and -0.605  1.512 and-1.302  Crystal Data  rn z Density (calculated)  (Mg/rn^)  Data Collection and Refinement  Goodness-of-fit on  *  Largest diff peak and hole (e" Á-^)  " R l on F = E I (|Fo| - |Fc|) | / S |Fo|, ( / > 2a(/)); wR2 = [ ( I ( Fo' -FÍf)li: data); w = [ CT'FO' ]"'; * GOF = [ E (W ( |Fo| - |Fc| f ) / deg?-ees offreedom  w(Fo' ff" f\  (aU  A.5 Examination of a Potential Catalytic Cycle for C-C and C-O Bond Formation Via C-H Bond Activation Dr. lan Blackmore and Chris Semiao investigated the possibility of a catalytic cycle for carbon-carbon and carbon-oxygen bond formation. Complex A26 in Scheme A. 12 represents a variation of the phosphine precatalysts. Complexes A27 and A28 are the products formed when the catalytic coupling is attempted with OC'Pri and benzene (Scheme A. 13). The catalytic cycle has been shown to be unfeasible, and the structures have been published."  Scheme A.12  A27  A28  Figure A.26 Solid-state molecular structure of A26 with 50 % probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-P(l) = 2.5087(11), W(l)-C(l) = 2.242(4), N(2)-C(l) = 1.452(5), P(l)-N(2) = 1.626(4), P(l)-N(3) = 1.680(3), P(l)-N(4) = 1.644(4), W(l)-N(l) = 1.766(4), N(l)-0(1) = 1.210(5), P(l)-W(l)-C(l) = 61.58(11), P(l)-N(2)C(l) = 104.9(3), W(l)-P(l)-N(2) = 88.73(13), N(2)-C(l)-W(l) = 104.4(3), W(l)-N(l)-0(1) = 168.3(4). Data for A26 were coUected to a máximum 26 valué of 55.6 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refmed anisotropically. A l l hydrogen atoms were included in fixed positions, except the hydride atom (HOl), the position of which was modeled based on residual electrón density in the appropriate open space. The final cycle of fiall-matrix least-squares analysis was based on 4580 observed reflecüons and 222 variable parameters. X-ray crystallographic data for the structure are presented in Table A. 11.  Figure A.27 Solid-state molecular structure of A27 with 50% probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-0(2) = 2.043(4), W(l)-C(l) = 2.211(6), 0(2)-C(l) = 1.388(7), W(l)-P(l) = 2.4863(16), W(l)-N(l) = 1.774(5), N(l)-0(1) = 1.222(6), 0(2)-C(l)-C(2) = 110.7(5), 0(2)-C(l)-C(5) = 110.9(5), C(2)-C(l)-C(5) = 120.5(6). W(l)-N(l)-0(1)= 173.6(5). The data for this structure were collected and initially solved by Brian Patrick. Refmements, including the disordered solvent, were done by the author. Data for A27 were collected to a máximum 29 valué of 50.0 ° in 0.5 ° oscillations. The structure was solved by direct methods and expanded using Fourier techniques. The crystal was a three-component twin. The unit cell contained a disordered THF solvent molecule, which was modeled in two orientations using isotropic carbón atoms. A l l other non-hydrogen atoms were refined anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of fullmatrix least-squares analysis was based on 5127 observed reflections and 415 variable parameters. X-ray crystallographic data for the structure are presented in Table A . l 1.  Figure A.28 Solid-state molecular structure of A28 with 50% probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-0(2) = 1.899(2), 0(2)-C(l) = 1.433(4), W(l)-C(8) = 2.142(3), W(l)-N(l) = 1.759(3), N(l)-0(1) = 1.227(3), W(l)-0(2)-C(l) = 135.7(2), 0(2)-C(l)-C(2) = 106.4(3), 0(2)-C(l)-C(5) =111.0(3), C(2)-C(l)-C(5) = 114.8(3), W(l)-N(l)-0(1)= 170.2(3). Data for A28 were coUected to a máximum 29 valué of 56.0 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refmed anisotropically, and all hydrogen atoms were included in fixed positions. The final cycle of fuU-matrix least-squares analysis was based on 5384 observed reflections and 257 variable parameters. X-ray crystallographic data for the structure are presented in Table A . l l .  Table A . l l X-ray Crystallographic Data for Complexes A26, A27 and A28. A26 A27 A28 Crystal Data Empirical formula  C16H33N4OPW  C39H50NO3PW  C23H35NO2W  Crystal Habit, color  Prism, yellow  Prism, yellow  Irregular, red  Crystal size (mm)  0.90 x 0.55 x 0.15  0.30x0.15x0.09  0.2x0.1 x O . l  Crystal system  Monoclinic  Triclinic  Monoclinic  Space group  P2i/c  P-i  Volume (Á^)  1961.1(4)  1822.5(5)  2259.55(9)  a (A)  15.711(2)  11.323(2)  8.2752(2)  biA)  7.7716(8)  12.829(2)  18.3553(4)  CiA)  17.039(2)  13.410(2)  15.2157(4)  «O  90  89.345(5)  90  109.499(6)  76.520(4)  102.1320(10)  rC) z  90  74.479(4)  90  4  2  4  Density (calculated) ( M g W )  1.735  1.450  1.591  Absorption coefficient (mm"')  5.981  3.249  5.129  .^000  1016  808  1080  Measured Reflections: Total  26735  5127  35475  Measured Reflections: Unique  4580  5127  5384  Final R índices"  R l = 0.0295, wR2  R l = 0.0337, wR2  R l = 0.0243, V  = 0.0791  = 0.0863  = 0.0474  Goodness-of-fit onF^^  1.068  1.071  1.097  Largest diff peak and hole (e'  1.079 and -2.373  1.475 and -1.551  0.781 and -0.797  Data Collection and Refínement  " R l on F = E I (|Fo| - | F , | ) | / E |Fo|, (/ > 2(T(7)); WR2 = [ ( E ( F^^ - F ^ ' )' ) / E w(Fo' )'] data); w = [ crVo^ ]"'; * GOF = [ E (w (|Fo| - \F,\ f ) I degrees offreedom ]1/2  (all  A.6 Selective Ortho-Activation of Aryl C-H Bonds by Cp*W(NO)(CH2CMe3)2 Jenkins Tsang investigated the selective ortho-activation of aryl substrates by 1, where the aryl substrates have the formulae CeHsX or C6H4X2 and X = diphenylacetylene, methoxy, chlorine or fluorine (Scheme A.14). Complexes A29, A30, A31 and A32 have been published.  Figure A.29 Solid-state molecular structure of A29 with 50% probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.156(4), W(l)-C(15) = 2.118(4), W(l)-N(l) = 1.767(3), N(l)-0(1) = 1.225(5), C(7)-C(8) = 1.203(6), W(l)-C(l)-C(6) = 123.5(3), C(l)-C(6)-C(7) = 122.4(3), C(6)-C(7)-C(8) = 173.8(4), C(7)-C(8)-C(9) = 175.9(4), W(l)-N(l)-0(1)= 170.3(3). Data for A29 were coUected to a máximum 26» valué of 55.6 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refmed anisotropically; hydrogen atoms H15A and H15B were refmed isotropically with fixed bond distances, and all other hydrogen atoms were included in fixed positions. The final cycle of ñiU-matrix least-squares analysis was based on 5848 observed reflections and 305 variable parameters. X-ray crystallographic data for the structure are presented in Table A.12.  C4 Figure A.30 Solid-state molecular structure of A30 with 50% probability thermal ellipsoids. Selected interatomic distances (Á) and angles (deg): W(l)-C(l) = 2.149(3), W(l)-C(8) = 2.122(4), W ( l ) - N ( l ) = 1.765(3), N(l)-0(2) = 1.223(4), W(l)-C(l)-C(2) = 119.7(2), C(l)-C(2)0(1)= 114.8(3), W(l)-C(8)-C(9)= 129.7(2), W(l)-N(l)-0(2) = 171.6(3). Data for A30 were collected to a máximum 20 valué of 55.8 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H08A and H08B were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of ñill-matrix least-squares analysis was based on 4666 observed reflections and 252 variable parameters. X-ray crystallographic data for the structure are presented in Table A.12.  Figure A.31 Solid-state molecular structure of A31 with 50% probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.162(3), W(l)-C(7) = 2.101(3), W(l)-N(l) = 1.775(2), N(l)-0(1) = 1.228(3), W(l)-C(l)-C(2) = 121.5(2), C(l)-C(2)F(l) = 117.4(3), W(l)-C(7)-C(8) = 130.9(2), W(l)-N(l)-0(1) = 169.4(2). Data for A31 were collected to a máximum 16 valué of 55.8 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H07A and H07B were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of full-matrix least-squares analysis was based on 4883 observed reflections and 243 variable parameters. X-ray crystallographic data for the structure are presented in Table A. 12.  Table A.12 X-ray Crystallographic Data for Complexes A29, A30 and A31. A29 A30 A31 Crystal Data Empirical formula  C29H35NOW  C22H33NO3W  C21H30FNOW  Crystal Habit, color  Píate, purple  Prism, dark red  Irregular, red  Crystal size (mm)  0.20x0.15x0.05  0.50 x 0.50 x 0.10  0.30x0.15x0.09  Crystal system  Triclinic  Monoclinic  Triclinic  Space group  P-i  P2i/n  P-i  Volume (Á^)  1284.4(2)  2143.2(3)  1041.71(19)  a(Á)  8.8372(8)  10.6196(11)  8.2185(9)  biA)  9.1967(8)  18.1403(13)  9.2843(9)  CiA)  16.620(2)  11.2112(10)  14.5576(15)  aC)  75.253(4)  90  74.260(4)  Pi°)  79.602(4)  97.101(3)  89.287(4)  rC)  88.771(4)  90  77.311(4)  z  2  4  2  Density (calculated) (Mg/m^)  1.545  1.634  1.643  Absorption coefficient (mm"')  4.517  5.405  5.561  596  1048  508  Measured Reflections: Total  56192  18559  34992  Measured Reflections: Unique  5848  4666  4883  Final R índices"  R l = 0.0245, wR2  R l = 0.0356, wR2  R l = 0.0221, wR2  = 0.0686  = 0.0925  = 0.0586  1.212  1.096  1.092  1.282 and-1.263  4.605 and -3.493  1.963 and-1.918  Data CoUection and Refinement  Goodness-of-fit on  *  Largest diff peak and hole (e"  " R l on F = S I (|Fol - IFcl) | / 1 |Fo|, (/> 2a(/)); wR2 = [ (E (F,^ - F o ^ ) ' ) / S w(Fo' data); w = [ ci^Fo^ ]"'; * G O F = [ S (w (\F^\ - \F,\ )^) I degrees offreedom  .  )^](all  Figure A.32 Solid-state molecular structure of A32 with 50% probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.1657(17), W(l)-C(7) = 2.1002(17), W ( l ) - N ( l ) = 1.7686(14), N(l)-0(1) = 1.2280(19), W(l)-C(l)-C(2) = 120.79(13), C(l)-C(2)-F(l) = 116.70(15), W(l)-C(l)-C(6) = 127.40(13), C(l)-C(6)-F(2) = 118.67(16), W ( l ) C(7)-C(8)= 131.50(12), W(l)-N(l)-0(1)= 169.27(13). Data for A32 were collected to a máximum 26 valué of 56.0 ° in 0.5 ° oscillations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refined anisotropically; hydrogen atoms H7A and H7B were refined isotropically, and all other hydrogen atoms were included in fixed positions. The final cycle of firll-matrix least-squares analysis was based on 4972 observed reflections and 251 variable parameters. X-ray crystallographic data for the structure are presented in Table A. 13.  Figure A.33 Solid-state molecular structure of A33 with 50% probability thermal ellipsoids. Selected interatomic distances (A) and angles (deg): W(l)-C(l) = 2.175(6), W(l)-C(7) = 2.102(6), W(l)-N(l) = 1.769(5), N(l)-0(1) = 1.238(6), W(l)-C(l)-C(2) = 124.4(5), C(l)-C(2)Cl(l) = 118.9(5), W(l)-C(7)-C(8) = 131.8(4), W(l)-N(l)-0(1) = 169.7(5). Data for A33 were coUected to a máximum 20 valué of 55.8 ° in 0.5 ° osciUations. The structure was solved by direct methods* and expanded using Fourier techniques. A l l nonhydrogen atoms were refmed anisotropically, and all hydrogen atoms were included in fixed positions. The fínal cycle of fiíU-matrix least-squares analysis was based on 5035 observed reflections and 234 variable parameters. X-ray crystallographic data for the structure are presented in Table A . 13. The R valué of A33 is somewhat higher than expected. A residual q-peak near C5 and in the plañe of the aryl ring suggests a small amount of disorder of the chlorine atom to that position; however, this disorder could not be modeled.  Table A.13 X-ray Crystallographic Data for Complexes A32 and A33. A32 A33 Crystal Data Empirical forrrmla  C21H29F2NOW  C21H30CINOW  Crystal Habit, color  Rod, red  Prism, red  Crystal size (mm)  0.50 X 0.20 x 0.20  0.2 X 0.2 X 0.2  Crystal system  Triclinic  Triclinic  Space group  P-i  P.i  Volume (Á^)  1047.99(15)  1074.9(2)  a (A)  8.0777(6)  8.3451(9)  b(A)  9.4908(8)  9.3513(11)  c(A)  14.7275(13)  14.6198(16)  «(°)  72.039(3)  74.282(4)  pn rn z  89.092(2)  89.927(4)  77.726(2)  78.609(4)  2  2  Density (calculated) (Mg/m^)  1.690  1.643  Absorption coefficient (mm"')  5.537  5.505  •Fooo  524  524  Measured Reflections: Total  20467  23570  Measured Reflections: Unique  4972  5035  Final R índices"  R l = 0.0126, wR2  R l = 0.0410, w  = 0.0310  = 0.1029  1.052  1.059  0.691 and -0.496  4.031 and -1.196  Data Collection and Refinement  Goodness-of-fit on  *  Largest diff peak and hole (e'  " R l o n F = S I (|Fo| - |Fc|) | / S |Fo|, (/>2a(/)); wR2 = [ (S (F,^ -F^^f)/! data); w = [ aVo^ ]''; * GOF = [ I (w (iFo| - |Fc| f ) / degrees offreedom  w(Fo' ) ' ] " ' (all .  A.7 References (1)  Cromer, D. T.; Waber, J. T. International Tables for X-ray Crystallography; Kynoch Press: Birmingham, 1974; Vol. IV.  (2)  Ibers, J. A . ; Hamilton, W. C. Acta Crystallogr. 1964, i 7, 781 -782.  (3)  Creagh, D. C ; McAuley, W. J. International Tables of X-ray Crystallography; Kluwer Academic Publishers: Boston, 1992; Vol. C.  (4)  Creagh, D. C ; Hubbell, J. H. International Tables for X-ray Crystallography; Kluwer Academic Publishers: Boston, 1992; Vol. C.  (5)  CrystalClear: Versionl .3.5b20; Molecular Structure Corporation, 2002.  (6)  SHELXL97: Sheldrick, G. M . University of Gottingen, Germany, 1997.  (7)  Jin, X . ; Legzdins, P.; Buschhaus, M . S. A. J. Am. Chem. Soc. 2005,127, 6928-6929.  (8)  SIR92: Altomare, A . ; Cascarano, M . ; Giacovazzo, C ; Guagliardi, A. J. Appl. Cryst. 1993, 26, 343.  (9)  (a) Tsang, J. Y . K.; Buschhaus, M . S. A.; Legzdins, P. J. Am. Chem. Soc. 2007,129, 5372-5373. (b) Tsang, J. Y . K.; Buschhaus, M . S. A . ; Graham, P. M . ; Semiao, C. J.; Semproni, S. P.; Kim, S. J.; Legzdins, P. J. Am. Chem. Soc. 2008, 130, 3652-3663.  (10)  Tsang, J. Y . K.; Buschhaus, M . S. A.; Fujita-Takayama, C ; Patrick, B. O.; Legzdins, P. Organometallics, 2008,27, 1634-1644.  (11)  Blackmore, I. J.; Semiao, C. J.; Buschhaus, M . S. A.; Patrick, B. O.; Legzdins, P. Organometallics, 2007, 26, 4881-4889.  (12)  Tsang, J. Y . K.; Buschhaus, M . S. A.; Legzdins, P.; Patrick, B. O. Organometallics, 2006, 25, 4215-4225.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0061738/manifest

Comment

Related Items